Fix try/catch catches more than it should #1859

[nicowilliams]

Rename the FORK_OPT opcode to TRY_BEGIN, add a TRY_END opcode, and wrap errors when raising through a TRY_END so that they will not be caught by the matching TRY_BEGIN.

Now a try exp catch handler expression generates code like:

TRY_BEGIN handler
<exp>
TRY_END
JUMP past_handler
handler: <handler>
past_handler:
...

On backtrack through TRY_BEGIN it just backtracks.

If anything past the whole thing raises when produced a value, then the TRY_END will catch the error, wrap it in another, and backtrack. The TRY_BEGIN will see a wrapped error and then it will unwrap and re-raise the error.

If raises, then TRY_BEGIN will catch the error and jump to the handler, but the TRY_BEGIN will not stack_save() in that case, so on raise/backtrack the TRY_BEGIN will not execute again (nor will the TRY_END).

---

Resolves #1859, resolves #1885, resolves #2140, resolves #2220, resolves #2485, resolves #2073.

[nicowilliams]

This is a straight-forward cherry-pick of ee36e9a from my old dlopen branch. There were minimal conflicts. Let's see if it works (er, well, it's failing jqtest for me locally, so it needs more work).

[emanuele6]

It does not matter much, but, probably because of the way you have ./configure-ed your build locally, in parser.c, all the #line directives have the hardcoded full path to your parser.y/parser.c files (/home/nico/ws/jq/src/parser.y) instead of just src/parser.y; and instead of the boring #ifndef YY_YY_SRC_PARSER_H_INCLUDED, it now uses #ifndef YY_YY__HOME_NICO_WS_JQ_SRC_PARSER_H_INCLUDED. =D

[nicowilliams]

It does not matter much, but, probably because of the way you have ./configure-ed your build locally, in parser.c, all the #line directives have the hardcoded full path to your parser.y/parser.c files (/home/nico/ws/jq/src/parser.y) instead of just src/parser.y; and instead of the boring #ifndef YY_YY_SRC_PARSER_H_INCLUDED, it now uses #ifndef YY_YY__HOME_NICO_WS_JQ_SRC_PARSER_H_INCLUDED. =D

😆

That's pretty funny. It's probably because I used --srcdir=/home/nico/ws/jq, except using --srcdir=.. didn't help. Hmm.

[nicowilliams]

I think the issues above are fixed.

[itchyny]

Here's a list of issues that this PR fixes. Hope some examples from there will be included.

• try/catch catches more than it should #1859
• jq 1.6 different (incorrect?) behavior for tonumber? // tostring compared to jq 1.5 #1885
• |=.? returns empty #2140
• Try/catch in update assignment #2220
• Strange behaviour when using two piped try catch error #2485
• Error in any(generator;condition) #2073

[itchyny]

Here's also an example; jq -n 'first(.?,.?)' that might have been fixed by this PR.

[nicowilliams]

@itchyny nice finds! I'll check them out and add tests based on them.

[nicowilliams]

Here's also an example; jq -n 'first(.?,.?)' that might have been fixed by this PR.

Oh, that's ugly, but no, this PR wouldn't fix that. This PR is strictly about try/catch, and first/1 doesn't use try/catch. and yes, this fixes it.

[nicowilliams]

Looks like I might have cherry-picked the wrong commit. See #1885 (comment).

[nicowilliams]

@itchyny I think this is ready now.

[emanuele6]

LGTM

[emanuele6]

I merged fixing 2073 => #2073 in the commit message.