From ad09d1f176dbba13e35cf783caeae01c93b9c6df Mon Sep 17 00:00:00 2001 From: Rafael Vuijk Date: Tue, 4 Aug 2026 11:13:34 +0000 Subject: [PATCH] Write out the issue links in comments rather than the shorthand A comment saying #531 only resolves inside GitHub's own interface. Read from a checkout, from git log, or from an editor, it names neither the project nor anything reachable. On a pull request the shorthand is linked automatically, so it reads fine there; in a comment nothing links it. Seven of these, all in comments this branch's author wrote. The six that were already here -- #327 three times in InnerSimplifyTest, #254 and #311 in FractionSimplify, #170 in PatternTest -- are left alone; say the word and they can go the same way. Comments only. 4009 tests. --- .../Simplification/Patterns/Patterns.ExpandFactorize.cs | 3 ++- .../Simplification/Patterns/Patterns.Trigonometry.cs | 2 +- Sources/Tests/UnitTests/Calculus/IntegrationTest.cs | 3 ++- Sources/Tests/UnitTests/Common/AlreadyFixedIssuesTest.cs | 3 ++- Sources/Tests/UnitTests/Common/NumericsRegressionTest.cs | 6 ++++-- Sources/Tests/UnitTests/PatternsTest/MultipleAngleTest.cs | 3 ++- 6 files changed, 13 insertions(+), 7 deletions(-) diff --git a/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.ExpandFactorize.cs b/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.ExpandFactorize.cs index 028a7d6fb..3637ff616 100644 --- a/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.ExpandFactorize.cs +++ b/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.ExpandFactorize.cs @@ -71,7 +71,8 @@ Sumf or Minusf when CollectCommonFactors(x) is { } collected => collected, /// /// /// The pairwise rules only ever see two adjacent terms of the sum tree, so a sum of - /// four was left half-factored at a*(c + d) + b*c + b*d -- that is #531. Sums + /// four was left half-factored at a*(c + d) + b*c + b*d -- that is + /// https://github.com/asc-community/AngouriMath/issues/531. Sums /// of two are left to those rules, which are older and better tested. /// /// when no factor is shared, so the rule does not fire. diff --git a/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.Trigonometry.cs b/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.Trigonometry.cs index 803222ed5..74979b84c 100644 --- a/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.Trigonometry.cs +++ b/Sources/AngouriMath/Functions/Simplification/Patterns/Patterns.Trigonometry.cs @@ -94,7 +94,7 @@ internal static partial class Patterns /// ever reached for numeric angles that are multiples of pi. Nothing opened a /// symbolic sin(2x), so cos(2x) - (1 - 2sin(x)^2) did not reduce to /// zero and neither did (sin(2t)csc(t))^2/4 - cos(2t) - sin(t)^2, which is - /// #557. + /// https://github.com/asc-community/AngouriMath/issues/557. /// internal static Entity ExpandMultipleAngleRules(Entity x) => x switch { diff --git a/Sources/Tests/UnitTests/Calculus/IntegrationTest.cs b/Sources/Tests/UnitTests/Calculus/IntegrationTest.cs index f3f6ef6c2..cf3263622 100644 --- a/Sources/Tests/UnitTests/Calculus/IntegrationTest.cs +++ b/Sources/Tests/UnitTests/Calculus/IntegrationTest.cs @@ -335,7 +335,8 @@ public void Test2() var expr = MathS.Sin(x); // Quadrature over a symmetric interval leaves a residual around 1e-17, which // downcasting used to round away. It no longer does, because the same rounding - // was destroying every legitimate small number (see IssueRegressionTest, #602). + // was destroying every legitimate small number (see IssueRegressionTest, and + // https://github.com/asc-community/AngouriMath/issues/602). // Asserted the way Test1 and Test3 assert, against the method's own accuracy. Assert.True(MathS.Compute.DefiniteIntegral(expr, x, -1, 1).Abs() < 1e-15); } diff --git a/Sources/Tests/UnitTests/Common/AlreadyFixedIssuesTest.cs b/Sources/Tests/UnitTests/Common/AlreadyFixedIssuesTest.cs index d83884313..0832aa5c4 100644 --- a/Sources/Tests/UnitTests/Common/AlreadyFixedIssuesTest.cs +++ b/Sources/Tests/UnitTests/Common/AlreadyFixedIssuesTest.cs @@ -86,7 +86,8 @@ public void Issue415_IntervalsUnite() => // https://github.com/asc-community/AngouriMath/issues/550 // The reporter's small system did not solve. The 25x26 one from the same report - // is a question of scale and is tracked separately at #608. + // is a question of scale and is tracked separately at + // https://github.com/asc-community/AngouriMath/issues/608. [Fact] public void Issue550_SmallLinearSystemSolves() { diff --git a/Sources/Tests/UnitTests/Common/NumericsRegressionTest.cs b/Sources/Tests/UnitTests/Common/NumericsRegressionTest.cs index 24dabfea5..73ebf60cd 100644 --- a/Sources/Tests/UnitTests/Common/NumericsRegressionTest.cs +++ b/Sources/Tests/UnitTests/Common/NumericsRegressionTest.cs @@ -70,7 +70,8 @@ public void Issue584_NonTerminatingExpansionRoundTrips() $"round-trip of {text} gave {back}"); } - // Adjacent to #584: a zero integer part used to render as the empty string, + // Adjacent to https://github.com/asc-community/AngouriMath/issues/584: + // a zero integer part used to render as the empty string, // so ToBaseN(0.5m, 2) produced ".1" and ToBaseN(0m, 2) produced "". [Theory] [InlineData(0, 2, "0")] @@ -171,7 +172,8 @@ public void Issue602_ExplicitToleranceIsStillRespected() // https://github.com/asc-community/AngouriMath/issues/561 // The reporter's 4x4 matrix spans 1e5 to 1e50, so entries of its inverse fall // below 1e-16 and downcasting rounded them to zero. Fixed by the same change as - // #602; pinned here because the matrix path reaches it by its own route. + // https://github.com/asc-community/AngouriMath/issues/602; + // pinned here because the matrix path reaches it by its own route. [Fact] public void Issue561_LargeValuedMatrixInvertsExactly() { diff --git a/Sources/Tests/UnitTests/PatternsTest/MultipleAngleTest.cs b/Sources/Tests/UnitTests/PatternsTest/MultipleAngleTest.cs index eb2d13d0c..f31b0feff 100644 --- a/Sources/Tests/UnitTests/PatternsTest/MultipleAngleTest.cs +++ b/Sources/Tests/UnitTests/PatternsTest/MultipleAngleTest.cs @@ -30,7 +30,8 @@ public void IdentitiesReduceToZero(string input) => // The same identity written as cos(2x) - (2cos(x)^2 - 1) gets as far as // 1 - cos(x)^2 - sin(x)^2 and stops: sin^2 + cos^2 = 1 is matched as a pair, and // there the pair sits either side of a subtraction. That is the same - // pairwise-versus-flattened gap as #531, in a product of trigonometric terms + // pairwise-versus-flattened gap as + // https://github.com/asc-community/AngouriMath/issues/531, in a product of trigonometric terms // rather than a sum, and is not something opening the angle can reach. [Fact] public void PythagoreanPairAcrossASubtractionIsStillMissed() =>