lix/tests/unit/libexpr/value/print.cc

783 lines
18 KiB
C++
Raw Normal View History

#include "tests/libexpr.hh"
#include "value.hh"
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
#include "print.hh"
namespace nix {
using namespace testing;
struct ValuePrintingTests : LibExprTest
{
template<class... A>
void test(Value v, std::string_view expected, A... args)
{
std::stringstream out;
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
v.print(state, out, args...);
ASSERT_EQ(out.str(), expected);
}
};
TEST_F(ValuePrintingTests, tInt)
{
Value vInt;
vInt.mkInt(10);
test(vInt, "10");
}
TEST_F(ValuePrintingTests, tBool)
{
Value vBool;
vBool.mkBool(true);
test(vBool, "true");
}
TEST_F(ValuePrintingTests, tString)
{
Value vString;
vString.mkString("some-string");
test(vString, "\"some-string\"");
}
TEST_F(ValuePrintingTests, tPath)
{
Value vPath;
vPath.mkString("/foo");
test(vPath, "\"/foo\"");
}
TEST_F(ValuePrintingTests, tNull)
{
Value vNull;
vNull.mkNull();
test(vNull, "null");
}
TEST_F(ValuePrintingTests, tAttrs)
{
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.symbols.create("one"), &vOne);
builder.insert(state.symbols.create("two"), &vTwo);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
test(vAttrs, "{ one = 1; two = 2; }");
}
TEST_F(ValuePrintingTests, tList)
{
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
Value vList;
state.mkList(vList, 5);
vList.bigList.elems[0] = &vOne;
vList.bigList.elems[1] = &vTwo;
vList.bigList.size = 3;
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vList, "[ 1 2 «nullptr» ]");
}
TEST_F(ValuePrintingTests, vThunk)
{
Value vThunk;
ExprInt e(0);
vThunk.mkThunk(nullptr, e);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vThunk, "«thunk»");
}
TEST_F(ValuePrintingTests, vApp)
{
Value vApp;
vApp.mkApp(nullptr, nullptr);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vApp, "«thunk»");
}
TEST_F(ValuePrintingTests, vLambda)
{
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
Env env {
.up = nullptr,
.values = { }
};
use byte indexed locations for PosIdx we now keep not a table of all positions, but a table of all origins and their sizes. position indices are now direct pointers into the virtual concatenation of all parsed contents. this slightly reduces memory usage and time spent in the parser, at the cost of not being able to report positions if the total input size exceeds 4GiB. this limit is not unique to nix though, rustc and clang also limit their input to 4GiB (although at least clang refuses to process inputs that are larger, we will not). this new 4GiB limit probably will not cause any problems for quite a while, all of nixpkgs together is less than 100MiB in size and already needs over 700MiB of memory and multiple seconds just to parse. 4GiB worth of input will easily take multiple minutes and over 30GiB of memory without even evaluating anything. if problems *do* arise we can probably recover the old table-based system by adding some tracking to Pos::Origin (or increasing the size of PosIdx outright), but for time being this looks like more complexity than it's worth. since we now need to read the entire input again to determine the line/column of a position we'll make unsafeGetAttrPos slightly lazy: mostly the set it returns is only used to determine the file of origin of an attribute, not its exact location. the thunks do not add measurable runtime overhead. notably this change is necessary to allow changing the parser since apparently nothing supports nix's very idiosyncratic line ending choice of "anything goes", making it very hard to calculate line/column positions in the parser (while byte offsets are very easy). (cherry picked from commit 5d9fdab3de0ee17c71369ad05806b9ea06dfceda) Change-Id: Ie0b2430cb120c09097afa8c0101884d94f4bbf34
2024-01-29 05:19:23 +00:00
PosTable::Origin origin = state.positions.addOrigin(std::monostate(), 1);
auto posIdx = state.positions.add(origin, 0);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
auto body = ExprInt(0);
auto formals = Formals {};
ExprLambda eLambda(posIdx, createSymbol("a"), &formals, &body);
Value vLambda;
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
vLambda.mkLambda(&env, &eLambda);
test(vLambda, "«lambda @ «none»:1:1»");
eLambda.setName(createSymbol("puppy"));
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vLambda, "«lambda puppy @ «none»:1:1»");
}
TEST_F(ValuePrintingTests, vPrimOp)
{
Value vPrimOp;
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrimOp primOp{
.name = "puppy"
};
vPrimOp.mkPrimOp(&primOp);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vPrimOp, "«primop puppy»");
}
TEST_F(ValuePrintingTests, vPrimOpApp)
{
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrimOp primOp{
.name = "puppy"
};
Value vPrimOp;
vPrimOp.mkPrimOp(&primOp);
Value vPrimOpApp;
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
vPrimOpApp.mkPrimOpApp(&vPrimOp, nullptr);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vPrimOpApp, "«partially applied primop puppy»");
}
TEST_F(ValuePrintingTests, vExternal)
{
struct MyExternal : ExternalValueBase
{
public:
std::string showType() const override
{
return "";
}
std::string typeOf() const override
{
return "";
}
virtual std::ostream & print(std::ostream & str) const override
{
str << "testing-external!";
return str;
}
} myExternal;
Value vExternal;
vExternal.mkExternal(&myExternal);
test(vExternal, "testing-external!");
}
TEST_F(ValuePrintingTests, vFloat)
{
Value vFloat;
vFloat.mkFloat(2.0);
test(vFloat, "2");
}
TEST_F(ValuePrintingTests, vBlackhole)
{
Value vBlackhole;
vBlackhole.mkBlackhole();
test(vBlackhole, "«potential infinite recursion»");
}
TEST_F(ValuePrintingTests, depthAttrs)
{
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
BindingsBuilder builderEmpty(state, state.allocBindings(0));
Value vAttrsEmpty;
vAttrsEmpty.mkAttrs(builderEmpty.finish());
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.symbols.create("one"), &vOne);
builder.insert(state.symbols.create("two"), &vTwo);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
builder.insert(state.symbols.create("nested"), &vAttrsEmpty);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
BindingsBuilder builder2(state, state.allocBindings(10));
builder2.insert(state.symbols.create("one"), &vOne);
builder2.insert(state.symbols.create("two"), &vTwo);
builder2.insert(state.symbols.create("nested"), &vAttrs);
Value vNested;
vNested.mkAttrs(builder2.finish());
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vNested, "{ nested = { ... }; one = 1; two = 2; }", PrintOptions { .maxDepth = 1 });
test(vNested, "{ nested = { nested = { ... }; one = 1; two = 2; }; one = 1; two = 2; }", PrintOptions { .maxDepth = 2 });
test(vNested, "{ nested = { nested = { }; one = 1; two = 2; }; one = 1; two = 2; }", PrintOptions { .maxDepth = 3 });
test(vNested, "{ nested = { nested = { }; one = 1; two = 2; }; one = 1; two = 2; }", PrintOptions { .maxDepth = 4 });
}
TEST_F(ValuePrintingTests, depthList)
{
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.symbols.create("one"), &vOne);
builder.insert(state.symbols.create("two"), &vTwo);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
BindingsBuilder builder2(state, state.allocBindings(10));
builder2.insert(state.symbols.create("one"), &vOne);
builder2.insert(state.symbols.create("two"), &vTwo);
builder2.insert(state.symbols.create("nested"), &vAttrs);
Value vNested;
vNested.mkAttrs(builder2.finish());
Value vList;
state.mkList(vList, 5);
vList.bigList.elems[0] = &vOne;
vList.bigList.elems[1] = &vTwo;
vList.bigList.elems[2] = &vNested;
vList.bigList.size = 3;
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(vList, "[ 1 2 { ... } ]", PrintOptions { .maxDepth = 1 });
test(vList, "[ 1 2 { nested = { ... }; one = 1; two = 2; } ]", PrintOptions { .maxDepth = 2 });
test(vList, "[ 1 2 { nested = { one = 1; two = 2; }; one = 1; two = 2; } ]", PrintOptions { .maxDepth = 3 });
test(vList, "[ 1 2 { nested = { one = 1; two = 2; }; one = 1; two = 2; } ]", PrintOptions { .maxDepth = 4 });
test(vList, "[ 1 2 { nested = { one = 1; two = 2; }; one = 1; two = 2; } ]", PrintOptions { .maxDepth = 5 });
}
struct StringPrintingTests : LibExprTest
{
template<class... A>
void test(std::string_view literal, std::string_view expected, unsigned int maxLength, A... args)
{
Value v;
v.mkString(literal);
std::stringstream out;
printValue(state, out, v, PrintOptions {
.maxStringLength = maxLength
});
ASSERT_EQ(out.str(), expected);
}
};
TEST_F(StringPrintingTests, maxLengthTruncation)
{
test("abcdefghi", "\"abcdefghi\"", 10);
test("abcdefghij", "\"abcdefghij\"", 10);
test("abcdefghijk", "\"abcdefghij\" «1 byte elided»", 10);
test("abcdefghijkl", "\"abcdefghij\" «2 bytes elided»", 10);
test("abcdefghijklm", "\"abcdefghij\" «3 bytes elided»", 10);
}
// Check that printing an attrset shows 'important' attributes like `type`
// first, but only reorder the attrs when we have a maxAttrs budget.
TEST_F(ValuePrintingTests, attrsTypeFirst)
{
Value vType;
vType.mkString("puppy");
Value vApple;
vApple.mkString("apple");
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.symbols.create("type"), &vType);
builder.insert(state.symbols.create("apple"), &vApple);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
test(vAttrs,
"{ type = \"puppy\"; apple = \"apple\"; }",
PrintOptions {
.maxAttrs = 100
});
test(vAttrs,
"{ apple = \"apple\"; type = \"puppy\"; }",
PrintOptions { });
}
TEST_F(ValuePrintingTests, ansiColorsInt)
{
Value v;
v.mkInt(10);
test(v,
ANSI_CYAN "10" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsFloat)
{
Value v;
v.mkFloat(1.6);
test(v,
ANSI_CYAN "1.6" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsBool)
{
Value v;
v.mkBool(true);
test(v,
ANSI_CYAN "true" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsString)
{
Value v;
v.mkString("puppy");
test(v,
ANSI_MAGENTA "\"puppy\"" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsStringElided)
{
Value v;
v.mkString("puppy");
test(v,
ANSI_MAGENTA "\"pup\" " ANSI_FAINT "«2 bytes elided»" ANSI_NORMAL,
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrintOptions {
.ansiColors = true,
.maxStringLength = 3
});
}
TEST_F(ValuePrintingTests, ansiColorsPath)
{
Value v;
v.mkPath(state.rootPath(CanonPath("puppy")));
test(v,
ANSI_GREEN "/puppy" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsNull)
{
Value v;
v.mkNull();
test(v,
ANSI_CYAN "null" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsAttrs)
{
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.symbols.create("one"), &vOne);
builder.insert(state.symbols.create("two"), &vTwo);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
test(vAttrs,
"{ one = " ANSI_CYAN "1" ANSI_NORMAL "; two = " ANSI_CYAN "2" ANSI_NORMAL "; }",
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsDerivation)
{
Value vDerivation;
vDerivation.mkString("derivation");
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.sType, &vDerivation);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
test(vAttrs,
ANSI_GREEN "«derivation»" ANSI_NORMAL,
PrintOptions {
.ansiColors = true,
.force = true,
.derivationPaths = true
});
test(vAttrs,
"{ type = " ANSI_MAGENTA "\"derivation\"" ANSI_NORMAL "; }",
PrintOptions {
.ansiColors = true,
.force = true
});
}
TEST_F(ValuePrintingTests, ansiColorsError)
{
Value throw_ = state.getBuiltin("throw");
Value message;
message.mkString("uh oh!");
Value vError;
vError.mkApp(&throw_, &message);
test(vError,
ANSI_RED
"«error: uh oh!»"
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
ANSI_NORMAL,
PrintOptions {
.ansiColors = true,
.force = true,
});
}
TEST_F(ValuePrintingTests, ansiColorsDerivationError)
{
Value throw_ = state.getBuiltin("throw");
Value message;
message.mkString("uh oh!");
Value vError;
vError.mkApp(&throw_, &message);
Value vDerivation;
vDerivation.mkString("derivation");
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.sType, &vDerivation);
builder.insert(state.sDrvPath, &vError);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
test(vAttrs,
"{ drvPath = "
ANSI_RED
"«error: uh oh!»"
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
ANSI_NORMAL
"; type = "
ANSI_MAGENTA
"\"derivation\""
ANSI_NORMAL
"; }",
PrintOptions {
.ansiColors = true,
.force = true
});
test(vAttrs,
ANSI_RED
"«error: uh oh!»"
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
ANSI_NORMAL,
PrintOptions {
.ansiColors = true,
.force = true,
.derivationPaths = true,
});
}
TEST_F(ValuePrintingTests, ansiColorsAssert)
{
ExprVar eFalse(state.symbols.create("false"));
eFalse.bindVars(state, state.staticBaseEnv);
ExprInt eInt(1);
ExprAssert expr(noPos, &eFalse, &eInt);
Value v;
state.mkThunk_(v, expr);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(v,
ANSI_RED "«error: assertion 'false' failed»" ANSI_NORMAL,
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrintOptions {
.ansiColors = true,
.force = true
});
}
TEST_F(ValuePrintingTests, ansiColorsList)
{
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
Value vList;
state.mkList(vList, 5);
vList.bigList.elems[0] = &vOne;
vList.bigList.elems[1] = &vTwo;
vList.bigList.size = 3;
test(vList,
"[ " ANSI_CYAN "1" ANSI_NORMAL " " ANSI_CYAN "2" ANSI_NORMAL " " ANSI_MAGENTA "«nullptr»" ANSI_NORMAL " ]",
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsLambda)
{
Env env {
.up = nullptr,
.values = { }
};
use byte indexed locations for PosIdx we now keep not a table of all positions, but a table of all origins and their sizes. position indices are now direct pointers into the virtual concatenation of all parsed contents. this slightly reduces memory usage and time spent in the parser, at the cost of not being able to report positions if the total input size exceeds 4GiB. this limit is not unique to nix though, rustc and clang also limit their input to 4GiB (although at least clang refuses to process inputs that are larger, we will not). this new 4GiB limit probably will not cause any problems for quite a while, all of nixpkgs together is less than 100MiB in size and already needs over 700MiB of memory and multiple seconds just to parse. 4GiB worth of input will easily take multiple minutes and over 30GiB of memory without even evaluating anything. if problems *do* arise we can probably recover the old table-based system by adding some tracking to Pos::Origin (or increasing the size of PosIdx outright), but for time being this looks like more complexity than it's worth. since we now need to read the entire input again to determine the line/column of a position we'll make unsafeGetAttrPos slightly lazy: mostly the set it returns is only used to determine the file of origin of an attribute, not its exact location. the thunks do not add measurable runtime overhead. notably this change is necessary to allow changing the parser since apparently nothing supports nix's very idiosyncratic line ending choice of "anything goes", making it very hard to calculate line/column positions in the parser (while byte offsets are very easy). (cherry picked from commit 5d9fdab3de0ee17c71369ad05806b9ea06dfceda) Change-Id: Ie0b2430cb120c09097afa8c0101884d94f4bbf34
2024-01-29 05:19:23 +00:00
PosTable::Origin origin = state.positions.addOrigin(std::monostate(), 1);
auto posIdx = state.positions.add(origin, 0);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
auto body = ExprInt(0);
auto formals = Formals {};
ExprLambda eLambda(posIdx, createSymbol("a"), &formals, &body);
Value vLambda;
vLambda.mkLambda(&env, &eLambda);
test(vLambda,
ANSI_BLUE "«lambda @ «none»:1:1»" ANSI_NORMAL,
PrintOptions {
.ansiColors = true,
.force = true
});
eLambda.setName(createSymbol("puppy"));
test(vLambda,
ANSI_BLUE "«lambda puppy @ «none»:1:1»" ANSI_NORMAL,
PrintOptions {
.ansiColors = true,
.force = true
});
}
TEST_F(ValuePrintingTests, ansiColorsPrimOp)
{
PrimOp primOp{
.name = "puppy"
};
Value v;
v.mkPrimOp(&primOp);
test(v,
ANSI_BLUE "«primop puppy»" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsPrimOpApp)
{
PrimOp primOp{
.name = "puppy"
};
Value vPrimOp;
vPrimOp.mkPrimOp(&primOp);
Value v;
v.mkPrimOpApp(&vPrimOp, nullptr);
test(v,
ANSI_BLUE "«partially applied primop puppy»" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsThunk)
{
Value v;
ExprInt e(0);
v.mkThunk(nullptr, e);
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
test(v,
ANSI_MAGENTA "«thunk»" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsBlackhole)
{
Value v;
v.mkBlackhole();
test(v,
ANSI_RED "«potential infinite recursion»" ANSI_NORMAL,
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsAttrsRepeated)
{
BindingsBuilder emptyBuilder(state, state.allocBindings(1));
Value vEmpty;
vEmpty.mkAttrs(emptyBuilder.finish());
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.symbols.create("a"), &vEmpty);
builder.insert(state.symbols.create("b"), &vEmpty);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
test(vAttrs,
"{ a = { }; b = " ANSI_MAGENTA "«repeated»" ANSI_NORMAL "; }",
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, ansiColorsListRepeated)
{
BindingsBuilder emptyBuilder(state, state.allocBindings(1));
Value vEmpty;
vEmpty.mkAttrs(emptyBuilder.finish());
Value vList;
state.mkList(vList, 3);
vList.bigList.elems[0] = &vEmpty;
vList.bigList.elems[1] = &vEmpty;
vList.bigList.size = 2;
test(vList,
"[ { } " ANSI_MAGENTA "«repeated»" ANSI_NORMAL " ]",
PrintOptions {
.ansiColors = true
});
}
TEST_F(ValuePrintingTests, listRepeated)
{
BindingsBuilder emptyBuilder(state, state.allocBindings(1));
Value vEmpty;
vEmpty.mkAttrs(emptyBuilder.finish());
Value vList;
state.mkList(vList, 3);
vList.bigList.elems[0] = &vEmpty;
vList.bigList.elems[1] = &vEmpty;
vList.bigList.size = 2;
test(vList, "[ { } «repeated» ]", PrintOptions { });
test(vList,
"[ { } { } ]",
PrintOptions {
.trackRepeated = false
});
}
TEST_F(ValuePrintingTests, ansiColorsAttrsElided)
{
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
BindingsBuilder builder(state, state.allocBindings(10));
builder.insert(state.symbols.create("one"), &vOne);
builder.insert(state.symbols.create("two"), &vTwo);
Value vAttrs;
vAttrs.mkAttrs(builder.finish());
test(vAttrs,
"{ one = " ANSI_CYAN "1" ANSI_NORMAL "; " ANSI_FAINT "«1 attribute elided»" ANSI_NORMAL " }",
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrintOptions {
.ansiColors = true,
.maxAttrs = 1
});
Value vThree;
vThree.mkInt(3);
builder.insert(state.symbols.create("three"), &vThree);
vAttrs.mkAttrs(builder.finish());
test(vAttrs,
"{ one = " ANSI_CYAN "1" ANSI_NORMAL "; " ANSI_FAINT "«2 attributes elided»" ANSI_NORMAL " }",
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrintOptions {
.ansiColors = true,
.maxAttrs = 1
});
}
TEST_F(ValuePrintingTests, ansiColorsListElided)
{
BindingsBuilder emptyBuilder(state, state.allocBindings(1));
Value vOne;
vOne.mkInt(1);
Value vTwo;
vTwo.mkInt(2);
Value vList;
state.mkList(vList, 4);
vList.bigList.elems[0] = &vOne;
vList.bigList.elems[1] = &vTwo;
vList.bigList.size = 2;
test(vList,
"[ " ANSI_CYAN "1" ANSI_NORMAL " " ANSI_FAINT "«1 item elided»" ANSI_NORMAL " ]",
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrintOptions {
.ansiColors = true,
.maxListItems = 1
});
Value vThree;
vThree.mkInt(3);
vList.bigList.elems[2] = &vThree;
vList.bigList.size = 3;
test(vList,
"[ " ANSI_CYAN "1" ANSI_NORMAL " " ANSI_FAINT "«2 items elided»" ANSI_NORMAL " ]",
Unify and refactor value printing Previously, there were two mostly-identical value printers -- one in `libexpr/eval.cc` (which didn't force values) and one in `libcmd/repl.cc` (which did force values and also printed ANSI color codes). This PR unifies both of these printers into `print.cc` and provides a `PrintOptions` struct for controlling the output, which allows for toggling whether values are forced, whether repeated values are tracked, and whether ANSI color codes are displayed. Additionally, `PrintOptions` allows tuning the maximum number of attributes, list items, and bytes in a string that will be displayed; this makes it ideal for contexts where printing too much output (e.g. all of Nixpkgs) is distracting. (As requested by @roberth in https://github.com/NixOS/nix/pull/9554#issuecomment-1845095735) Please read the tests for example output. Future work: - It would be nice to provide this function as a builtin, perhaps `builtins.toStringDebug` -- a printing function that never fails would be useful when debugging Nix code. - It would be nice to support customizing `PrintOptions` members on the command line, e.g. `--option to-string-max-attrs 1000`. (cherry picked from commit 0fa08b451682fb3311fe58112ff05c4fe5bee3a4, ) === Restore ambiguous value printer for `nix-instantiate` The Nix team has requested that this output format remain unchanged. I've added a warning to the man page explaining that `nix-instantiate --eval` output will not parse correctly in many situations. (cherry picked from commit df84dd4d8dd3fd6381ac2ca3064432ab31a16b79) Change-Id: I7cca6b4b53cd0642f2d49af657d5676a8554c9f8
2024-03-08 02:05:47 +00:00
PrintOptions {
.ansiColors = true,
.maxListItems = 1
});
}
} // namespace nix