qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Eric Blake <eblake@redhat.com>
To: Markus Armbruster <armbru@redhat.com>
Cc: marcandre.lureau@redhat.com, qemu-devel@nongnu.org,
	mdroth@linux.vnet.ibm.com
Subject: Re: [PATCH 6/7] qapi: Split up scripts/qapi/common.py
Date: Wed, 2 Oct 2019 10:27:38 -0500	[thread overview]
Message-ID: <70ff7c98-1775-ad6a-e505-8f0a97210379@redhat.com> (raw)
In-Reply-To: <87lfu3rxdv.fsf@dusky.pond.sub.org>

On 10/2/19 10:16 AM, Markus Armbruster wrote:
> Eric Blake <eblake@redhat.com> writes:
> 
>> On 10/1/19 2:15 PM, Markus Armbruster wrote:
>>> The QAPI code generator clocks in at some 3100 SLOC in 8 source files.
>>> Almost 60% of the code is in qapi/common.py.  Split it into more
>>> focused modules:
>>>
>>> * Move QAPISchemaPragma and QAPISourceInfo to qapi/source.py.
>>>
>>> * Move QAPIError and its sub-classes to qapi/error.py.
>>>
>>> * Move QAPISchemaParser and QAPIDoc to parser.py.  Use the opportunity
>>>     to put QAPISchemaParser first.
>>>
>>> * Move check_expr() & friends to qapi/expr.py.  Use the opportunity to
>>>     put the code into a more sensible order.
>>
>> Code motion can be easier to review when it is 1:1 (using 'diff -u
>> <(sed -n '/^-//p' patch) <(sed -n '/^\+//p'patch)', which is quite
>> small if code moved wholesale).  Reordering things breaks that
>> property.
> 
> True.  But see below.

> Perhaps a bit of shell wizardry can increase your confidence.
> 
> Before this patch:
> 
> 1. Split into classes and functions (crudely!):
> 
>      $ csplit scripts/qapi/common.py '/^\(class\|def\) /' '{*}'
> 
> 2. Rename the parts:
> 
>      $ for i in xx*; do n=`sed -n '1s/^[a-z]* \([A-Za-z0-9_]*\).*/\1/p' $i`; [ "$n" ] && mv $i xx-$n; done
> 
> 3. Stash them:
> 
>      $ mkdir o
>      $ $ mv xx* o
> 
> After this patch:
> 
> 1. Split:
> 
>      $ csplit <(cat scripts/qapi/{common,source,error,parser,expr,schema,gen}.py) '/^\(class\|def\) /' '{*}'
> 
> 2. Rename:
> 
>      $ for i in xx*; do n=`sed -n '1s/^[a-z]* \([A-Za-z0-9_]*\).*/\1/p' $i`; [ "$n" ] && mv $i xx-$n; done
> 
> 3. Stash & diff:
> 
>      $ mkdir n
>      $ mv xx* n
>      $ diff -rup o n

Slick.

> 
> Output of diff appended for your reading pleasure.
> 
>> Reviewed-by: Eric Blake <eblake@redhat.com>
> 
> Thanks!
> 
> 
> diff -rup o/xx-QAPIDoc n/xx-QAPIDoc
> --- o/xx-QAPIDoc	2019-10-02 17:02:35.984552694 +0200
> +++ n/xx-QAPIDoc	2019-10-02 17:06:17.930607336 +0200
> @@ -273,5 +273,31 @@ class QAPIDoc(object):
>                   self.info,
>                   "the following documented members are not in "
>                   "the declaration: %s" % ", ".join(bogus))
> +#
> +# Check (context-free) QAPI schema expression structure
> +#

New boilerplate is obviousl

> +
> +# Names must be letters, numbers, -, and _.  They must start with letter,
> +# except for downstream extensions which must start with __RFQDN_.
> +# Dots are only valid in the downstream extension prefix.
> +valid_name = re.compile(r'^(__[a-zA-Z0-9.-]+_)?'
> +                        '[a-zA-Z][a-zA-Z0-9_-]*$')
>   

The crude split shows that this blurb changed in relation to which 
class/def it was closest to (but you did say it was a crude split), 
which isn't fatal.

>   
> diff -rup o/xx-QAPIGen n/xx-QAPIGen
> --- o/xx-QAPIGen	2019-10-02 17:02:35.987552655 +0200
> +++ n/xx-QAPIGen	2019-10-02 17:06:17.932607309 +0200
> @@ -43,4 +43,3 @@ class QAPIGen(object):
>           f.close()
>   
>   
> -@contextmanager
> diff -rup o/xx-QAPIGenH n/xx-QAPIGenH
> --- o/xx-QAPIGenH	2019-10-02 17:02:35.987552655 +0200
> +++ n/xx-QAPIGenH	2019-10-02 17:06:17.933607296 +0200
> @@ -7,3 +7,4 @@ class QAPIGenH(QAPIGenC):
>           return guardend(self.fname)
>   
>   
> +@contextmanager

Another victim of the crude split (this line logically begins with the 
next line in the original file(s), not the previous class/def).

> diff -rup o/xx-QAPISchema n/xx-QAPISchema
> --- o/xx-QAPISchema	2019-10-02 17:02:35.986552668 +0200
> +++ n/xx-QAPISchema	2019-10-02 17:06:17.932607309 +0200
> @@ -297,9 +297,26 @@ class QAPISchema(object):
>                       visitor.visit_module(module)
>                   entity.visit(visitor)
>           visitor.visit_end()
> -
> -
>   #
> -# Code generation helpers
> +# QAPI code generation
> +#

More whitespace and boilerplate.


> +++ n/xx-QAPISchemaParser	2019-10-02 17:06:17.930607336 +0200
> @@ -268,14 +268,3 @@ class QAPISchemaParser(object):
>           raise QAPIParseError(self, "documentation comment must end with '##'")
>   
>   
> -#
> -# Check (context-free) schema expression structure
> -#
> -
> -# Names must be letters, numbers, -, and _.  They must start with letter,
> -# except for downstream extensions which must start with __RFQDN_.
> -# Dots are only valid in the downstream extension prefix.
> -valid_name = re.compile(r'^(__[a-zA-Z0-9.-]+_)?'
> -                        '[a-zA-Z][a-zA-Z0-9_-]*$')
> -
> -

We already visited this above.


>   
> Only in o: xx-camel_case

Interesting change, but not detrimental.

Overall, the interdiff is fairly representative of clean code motion in 
spite of any reordering.  Thanks for doing that!

R-b still stands.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org


  reply	other threads:[~2019-10-02 15:29 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-01 19:15 [PATCH 0/7] qapi: Cleanups and test speedup Markus Armbruster
2019-10-01 19:15 ` [PATCH 1/7] qapi: Don't suppress doc generation without pragma doc-required Markus Armbruster
2019-10-01 19:59   ` Eric Blake
2019-10-01 19:15 ` [PATCH 2/7] qapi: Store pragma state in QAPISourceInfo, not global state Markus Armbruster
2019-10-01 20:01   ` Eric Blake
2019-10-01 19:15 ` [PATCH 3/7] qapi: Eliminate accidental global frontend state Markus Armbruster
2019-10-01 20:03   ` Eric Blake
2019-10-01 19:15 ` [PATCH 4/7] qapi: Speed up frontend tests Markus Armbruster
2019-10-01 20:23   ` Eric Blake
2019-10-02 14:31     ` Markus Armbruster
2019-10-01 19:15 ` [PATCH 5/7] qapi: Move gen_enum(), gen_enum_lookup() back to qapi/types.py Markus Armbruster
2019-10-01 20:26   ` Eric Blake
2019-10-01 19:15 ` [PATCH 6/7] qapi: Split up scripts/qapi/common.py Markus Armbruster
2019-10-01 21:19   ` Eric Blake
2019-10-02 15:16     ` Markus Armbruster
2019-10-02 15:27       ` Eric Blake [this message]
2019-10-02 16:13         ` Markus Armbruster
2019-10-16 13:05   ` Kevin Wolf
2019-10-16 14:09     ` Markus Armbruster
2019-10-01 19:15 ` [PATCH 7/7] qapi: Clear scripts/qapi/doc.py executable bits again Markus Armbruster
2019-10-01 20:34   ` Eric Blake

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=70ff7c98-1775-ad6a-e505-8f0a97210379@redhat.com \
    --to=eblake@redhat.com \
    --cc=armbru@redhat.com \
    --cc=marcandre.lureau@redhat.com \
    --cc=mdroth@linux.vnet.ibm.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).