Bug 10859 – Problem with array() of a map Range of simple tuples

I don't know if this is a regression. Here "words.txt" is just a file of words separated by a newline: import std.stdio: File; import std.algorithm: map; import std.typecons: tuple; import std.array: array; void main() { File("words.txt") .byLine .map!(line => tuple(line)) .array ; } If I compile with -g it gives at run-time, dmd 2.064alpha: [email protected](1927): Assertion failure ---------------- 0x00424903 in onAssertError 0x0040ED86 in pure nothrow @safe void std.typecons.Tuple!(char[]).Tuple.opAssign!(std.typecons.Tuple!(char[]).Tuple).opAssign(std.typecons.Tuple!(char[]).Tuple) at ...\dmd2\src\phobos\std\typecons.d(525) 0x0040F5A4 in pure nothrow @safe void std.array.Appender!(std.typecons.Tuple!(char[]).Tuple[]).Appender.put!(std.typecons.Tuple!(char[]).Tuple).put(std.typecons.Tuple!(char[]).Tuple) at ...\dmd2\src\phobos\std\array.d(2347) 0x0040EE72 in D3std5array116__T5arrayTS6test5b4main80__T9MapResultS226test5b4main9__lambda2TS3std5stdio4File1F8E0F5CC6B0162C488BCADE9A33A8B01 at ...\dmd2\src\phobos\std\array.d(70) 0x00402087 in _Dmain at ...\test5b.d(7) 0x0041C580 in extern (C) int rt.dmain2._d_run_main(int, char**, extern (C) int function(char[][])*).void runMain() 0x0041C610 in extern (C) int rt.dmain2._d_run_main(int, char**, extern (C) int function(char[][])*).void runAll() 0x0041BEED in _d_run_main 0x004084D0 in main 0x00432825 in mainCRTStartup 0x7691D2E9 in BaseThreadInitThunk 0x76FB1603 in RtlInitializeExceptionChain 0x76FB15D6 in RtlInitializeExceptionChain ---------------- If inside the main I replace "array" with "writeln" the program works correctly.

Kind of like #10690: Combination bug Array calls appender. appender calls opAssign on un-initilized memory (which is wrong: It should call emplace... but emplace is broken, so that wouldn't change much anyways) In any case, once inside Tuple.opAssign, the implementation will *swap* (correctly). Unfortunatly, since "this" is not initialized, an assert will trigger in "swap", as the string "points" to the other object. There are many steps involved for a "full" recovery, but I think that removing the "pointsTo" assertion in "swap" is the first step: http://d.puremagic.com/issues/show_bug.cgi?id=9975#c14 https://github.com/D-Programming-Language/phobos/pull/1390

It seems that http://d.puremagic.com/issues/show_bug.cgi?id=10136 is probably related to this issue.

(In reply to comment #2) > It seems that http://d.puremagic.com/issues/show_bug.cgi?id=10136 is probably > related to this issue. It seems Issue 10136 is a worksforme.

(In reply to comment #1) > Kind of like #10690: Combination bug > > Array calls appender. > > appender calls opAssign on un-initilized memory (which is wrong: It should call > emplace... but emplace is broken, so that wouldn't change much anyways) > > In any case, once inside Tuple.opAssign, the implementation will *swap* > (correctly). > > Unfortunatly, since "this" is not initialized, an assert will trigger in > "swap", as the string "points" to the other object. > > There are many steps involved for a "full" recovery, but I think that removing > the "pointsTo" assertion in "swap" is the first step: > http://d.puremagic.com/issues/show_bug.cgi?id=9975#c14 > https://github.com/D-Programming-Language/phobos/pull/1390 Still same symptoms as 10690: Not fixed because appender calls opAssign.

(In reply to comment #4) > (In reply to comment #1) > > Kind of like #10690: Combination bug > > > > Array calls appender. > > > > appender calls opAssign on un-initilized memory (which is wrong: It should call > > emplace... but emplace is broken, so that wouldn't change much anyways) > > > > In any case, once inside Tuple.opAssign, the implementation will *swap* > > (correctly). > > > > Unfortunatly, since "this" is not initialized, an assert will trigger in > > "swap", as the string "points" to the other object. > > > > There are many steps involved for a "full" recovery, but I think that removing > > the "pointsTo" assertion in "swap" is the first step: > > http://d.puremagic.com/issues/show_bug.cgi?id=9975#c14 > > https://github.com/D-Programming-Language/phobos/pull/1390 > > Still same symptoms as 10690: Not fixed because appender calls opAssign. https://github.com/D-Programming-Language/phobos/pull/1529

This is a clear dup caused by same appender issue (onAssign call on garbage). *** This issue has been marked as a duplicate of issue 10690 ***