Zsh Mailing List Archive
Messages sorted by:
Reverse Date,
Date,
Thread,
Author
Re: vcs_info: '%' in payloads not escaped
- X-seq: zsh-workers 40398
- From: Daniel Shahaf <d.s@xxxxxxxxxxxxxxxxxx>
- To: zsh-workers@xxxxxxx
- Subject: Re: vcs_info: '%' in payloads not escaped
- Date: Mon, 23 Jan 2017 11:04:32 +0000
- Dkim-signature: v=1; a=rsa-sha1; c=relaxed/relaxed; d= daniel.shahaf.name; h=content-transfer-encoding:content-type :date:from:in-reply-to:message-id:mime-version:references :subject:to:x-me-sender:x-me-sender:x-sasl-enc:x-sasl-enc; s= mesmtp; bh=xAO/zpx3P1Qv3BPhOOkA/ouAv7Q=; b=kypk4tQIeKhZM7XI0WW3r 7feB/u3FCQ9MObs6gcpzpufyTEIOtGMCBvKhx1C+HtSPZ+FTV0Wpry0pek9Qa/kI cnSN3vXQsunpgMVF1SOIcs04FavZG1CQlPz5KB9YGRLHvzKzYoOPk7oPo7iqx+FW SsVcDUup5b2QbSE/dJgDCU=
- Dkim-signature: v=1; a=rsa-sha1; c=relaxed/relaxed; d= messagingengine.com; h=content-transfer-encoding:content-type :date:from:in-reply-to:message-id:mime-version:references :subject:to:x-me-sender:x-me-sender:x-sasl-enc:x-sasl-enc; s= smtpout; bh=xAO/zpx3P1Qv3BPhOOkA/ouAv7Q=; b=d7CyNu4aVk9rvDVqCbHA Fje7ukn04Rsnd5lPo/kGo/UYQ/ceYIJp4tCOKrIpAro2oCkEwccR/jRjwBvNrp3J GPGYEAaKmOot0nTUxTxErAY9QH0P+YBLI6fjPsFv0XBSXAXIqsaUa+0pb6CCfD8U DmW0PkDxXzfECOuJcsxwpag=
- In-reply-to: <170106092711.ZM7535@torch.brasslantern.com>
- List-help: <mailto:zsh-workers-help@zsh.org>
- List-id: Zsh Workers List <zsh-workers.zsh.org>
- List-post: <mailto:zsh-workers@zsh.org>
- Mailing-list: contact zsh-workers-help@xxxxxxx; run by ezmlm
- References: <20161227150507.GA20351@fujitsu.shahaf.local2> <20170105160730.GA21106@fujitsu.shahaf.local2> <871swhqxph.fsf@ft.bewatermyfriend.org> <20170106022128.GA6197@fujitsu.shahaf.local2> <87wpe8pj0o.fsf@ft.bewatermyfriend.org> <20170106164012.GB4948@fujitsu.shahaf.local2> <170106092711.ZM7535@torch.brasslantern.com>
Coming back to this thread...
Bart Schaefer wrote on Fri, Jan 06, 2017 at 09:27:11 -0800:
> I'm not very familiar with VCS_info. Frank seemed concerned that the
> same variable might contain substrings both from git output that
> had one set of quoting needs and also from other sources that had
> another set of quoting needs.
That's exactly what happens.
The code flow is:
1 if gen-applied-string hook is active:
2 applied_string=$(what the hook returns)
3 else:
4 applied_string="literal value from the VCS"
5
6 if set-patch-format hook is active:
7 hook_com[applied]=$applied_string
8 percent_m=$(what the hook returns)
9 else:
10 percent_m=$(zformat ... p:${applied_string})
11
12 VCS_INFO_formats … $percent_m
The catch is that both hooks have access to $applied_string, and either
of them might %-escape it and added prompt sequences (e.g., for
coloring) around it. That is: once gen-applied-string has run, we don't
know whether it has already %-escaped the string it returns, and by
extension we don't know whether set-patch-format expects its
${hook_com[applied]} input to be literal or %-escaped.
So perhaps what we _can_ do is:
1) If *neither* hook was called, have vcs_info itself perform %-escaping
appropriately. (That is: perform escaping on line 7 iff line 4 was
taken.)
2) Document that if *either* hook is registered, then handling
%-escaping is up to the user.
Would that address everyone's concerns?
Cheers,
Daniel
(optimizing for backwards compatibility)
Messages sorted by:
Reverse Date,
Date,
Thread,
Author