[Mesa-dev] [PATCH 5/6] vulkan: enum generator: Stop using iterparse

[Dylan Baker]

Quoting Jason Ekstrand (2017-09-21 08:32:22)
> While using iterparse is potentially a little more efficient, the Vulkan
> registry XML is not large and using regular element tree simplifies the
> parsing logic substantially.
> ---
>  src/vulkan/util/gen_enum_to_str.py | 41 +++++++++++++++-----------------------
>  1 file changed, 16 insertions(+), 25 deletions(-)
> 
> diff --git a/src/vulkan/util/gen_enum_to_str.py b/src/vulkan/util/gen_enum_to_str.py
> index 8f32102..e5f8964 100644
> --- a/src/vulkan/util/gen_enum_to_str.py
> +++ b/src/vulkan/util/gen_enum_to_str.py
> @@ -157,31 +157,22 @@ def parse_xml(enum_factory, ext_factory, filename):
>      of VkEnum objects.
>      """
>  
> -    with open(filename, 'rb') as f:
> -        context = iter(et.iterparse(f, events=('start', 'end')))
> -
> -        # This gives the root element, since goal is to iterate over the
> -        # elements without building a tree, this allows the root to be cleared
> -        # (erase the elements) after the children have been processed.
> -        _, root = next(context)
> -
> -        for event, elem in context:
> -            if event == 'end' and elem.tag == 'enums':
> -                type_ = elem.attrib.get('type')
> -                if type_ == 'enum':
> -                    enum = enum_factory(elem.attrib['name'])
> -                    enum.values.extend([e.attrib['name'] for e in elem
> -                                        if e.tag == 'enum'])
> -            elif event == 'start' and elem.tag == 'extension':
> -                ext_factory(elem.attrib['name'], int(elem.attrib['number']))
> -            elif event == 'end' and elem.tag == 'extension':
> -                if elem.attrib['supported'] != 'vulkan':
> -                    continue
> -                for e in elem.findall('.//enum[@extends][@offset]'):
> -                    enum = enum_factory(e.attrib['extends'])
> -                    enum.values.append(e.attrib['name'],)
> -
> -            root.clear()
> +    xml = et.parse(filename)
> +
> +    for enum_type in xml.findall('./enums'):
> +        if enum_type.attrib.get('type') != 'enum':
> +            continue

please do not use continue for iterating xlm. xpath is more than capable of
representing this in less code, and it's implemented in C so it will be much
faster.

for enum_type in xml.findall('./enums[@type="enum"]')

> +
> +        enum = enum_factory(enum_type.attrib['name'])
> +        for value in enum_type.findall('./enum'):
> +            enum.values.append(value.attrib['name'])
> +
> +    for ext_elem in xml.findall('./extension'):
> +        if ext_elem.attrib['supported'] != 'vulkan':
> +            continue

please do the same thing here.

> +
> +        ext_factory(ext_elem.attrib['name'],
> +                    int(ext_elem.attrib['number']))
>  
>  def main():
>      parser = argparse.ArgumentParser()
> -- 
> 2.5.0.400.gff86faf
> 
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/mesa-dev
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: signature
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20170921/06b751ce/attachment.sig>