[v2] Script to generate ChangeLog-like output

Message ID 20190112112444.26801-1-siddhesh@sourceware.org
State Committed
Headers

Commit Message

Siddhesh Poyarekar Jan. 12, 2019, 11:24 a.m. UTC
  The utility of a ChangeLog file has been discussed in various mailing
list threads and GNU Tools Cauldrons in the past years and the general
consensus is that while the file may have been very useful in the past
when revision control did not exist or was not as powerful as it is
today, it's current utility is fast diminishing.  Further, the
ChangeLog format gets in the way of modernisation of processes since
it almost always results in rewriting of a commit, thus preventing use
of any code review tools to automatically manage patches in the glibc
project.

There is consensus in the glibc community that documentation of why a
change was done (i.e. a detailed description in a git commit) is more
useful than what changed (i.e. a ChangeLog entry) since the latter can
be deduced from the patch.  The GNU community would however like to
keep the option of ascertaining what changed through a ChangeLog-like
output and as a compromise, it was proposed that a script be developed
that generates this output.

The script below is the result of these discussions.  This script
takes two git revisions references as input and generates the git log
between those revisions in a form that resembles a ChangeLog.  Its
capabilities and limitations are listed in a comment in the script.
On a high level it is capable of parsing C code and telling what
changed at the top level, but not within constructs such as functions.

For input other than C, the script only identifies if a file has been
added, removed, modified, permissions changed, etc. but cannot
understand the change in content.  The design of the script however is
pluggable, so it should be possible to develop additional parsers to
process other types of files.

I have tested it with a number of commits in the glibc log and also
fixed a couple of errors that were reported earlier.

Transition:

Once this script is in place, it should be possible for us to stop
maintaining the ChangeLog file and rely on the ChangeLog script to
give an output that serves a similar purpose.  Given that the majority
of our code is in C, we have adequate coverage with just the C parser.
In any case the readability of ChangeLog entries for other formats
(makefiles for example) is just too convoluted and is perhaps not even
worth the effort.

I propose that we stop ChangeLog file maintenance once 2.30 opens for
development in February and focus on making the ChangeLog script more
accurate if we encounter bugs.  I will also take another swipe at
patchwork to try and automate things in it now that the ChangeLog is
gone.

If there is agreement, then as part of 2.29 release management I will
update the wiki to reflect the change in our patch submission process
and also mention the Changelog script there for those who need it.  I
believe Joseph is working with RMS to change the wording in the GNU
Coding Standards to make ChangeLog management optional.

Looking forward, once 2.30 is released in August we will be in a good
position to decide if patchwork is useful or if we want to consider
other alternative patch review processes and tools.

ChangeLog:

	* scripts/gen-changed-entities.py: New script.
---

Changes from v1:

- Rewrote the macro nesting detection logic.
- Changed the way macro hacks are detected, now I build a list of macro
  definitions in addition to reading libc-symbols.h for symbol hack
  definitions.  This in combination with manual substitutions for known
  bad macros seems sufficient for accurate parsing.
- Fixed multiple issues reported by Richard and Alfred.

 scripts/gen-changed-entities.py | 1099 +++++++++++++++++++++++++++++++
 1 file changed, 1099 insertions(+)
 create mode 100755 scripts/gen-changed-entities.py
  

Comments

Richard Stallman Jan. 13, 2019, 9:34 p.m. UTC | #1
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

Thank you very much for developing this script.
But we can't presume it works entirely correctly yet.
Programs tend to have bugs, until we fix them.

So I ask people to test this script on a large number
of actual past commits, and compare the list of entities
reported by the script with what we have in the ChangeLog files
and with a hand analysis of the actual diff.  We will most likely
find some problems.

With attention of this kind, we can have it working solidly
to carry out your plan.
  
Siddhesh Poyarekar Jan. 14, 2019, 2:38 p.m. UTC | #2
On 14/01/19 3:04 AM, Richard Stallman wrote:
> [[[ To any NSA and FBI agents reading my email: please consider    ]]]
> [[[ whether defending the US Constitution against all enemies,     ]]]
> [[[ foreign or domestic, requires you to follow Snowden's example. ]]]
>
> Thank you very much for developing this script.
> But we can't presume it works entirely correctly yet.
> Programs tend to have bugs, until we fix them.


Thanks, the plan is to actively maintain the script to add features all 
the way into 2.30 and maybe even 2.31.  The idea of stopping ChangeLog 
writing in 2.29 is to allow us to clean up patchwork and start cleaning 
up our own workflows so that we can automate our patch reviews (or at 
least begin the process) by 2.30 release.  That way by 2.31 we will have 
the automation in place.


It is of course a pie-in-the-sky kind of plan but hopefully we will be 
able to pull it off.  As for the script, I have added it is a project in 
our local makerspace (I get requests from students for projects, 
hopefully this is a simple enough start) and if it is considered 
important enough for GSoC, I don't mind mentoring students working 
globally on additional backends for the script.


Siddhesh
  
Siddhesh Poyarekar Jan. 14, 2019, 2:43 p.m. UTC | #3
On 14/01/19 8:08 PM, Siddhesh Poyarekar wrote:
> Thanks, the plan is to actively maintain the script to add features all 
> the way into 2.30 and maybe even 2.31.  The idea of stopping ChangeLog 
> writing in 2.29 is to allow us to clean up patchwork and start cleaning 
> up our own workflows so that we can automate our patch reviews (or at 
> least begin the process) by 2.30 release.  That way by 2.31 we will have 
> the automation in place.
> 
> 
> It is of course a pie-in-the-sky kind of plan but hopefully we will be 
> able to pull it off.  As for the script, I have added it is a project in 
> our local makerspace (I get requests from students for projects, 
> hopefully this is a simple enough start) and if it is considered 
> important enough for GSoC, I don't mind mentoring students working 
> globally on additional backends for the script.

I've also thought about some other aspects of this:

1. Should this be a separate project that does only the job of 
generating a ChangeLog from any git log?

2. Should this be rewritten in C/C++ to make it significantly faster? 
Probably not a priority, but maybe something worth thinking about?

This of course is orthogonal to the glibc question, which this script 
should be capable of solving since a significant chunk is C code.

Siddhesh
  
Richard Stallman Jan. 15, 2019, 9:36 p.m. UTC | #4
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > Thanks, the plan is to actively maintain the script to add features all 
  > the way into 2.30 and maybe even 2.31.

Thank you -- that's doing it right.
  
Richard Stallman Jan. 15, 2019, 9:36 p.m. UTC | #5
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > 1. Should this be a separate project that does only the job of 
  > generating a ChangeLog from any git log?

I think so.  It does not belong in any other GNU package as far as I
know.  I am sure maintainers of various GNU packages will want to use
it.

  > 2. Should this be rewritten in C/C++ to make it significantly faster? 
  > Probably not a priority, but maybe something worth thinking about?

First let's see if it is fast enough.  If not, we can redo it later.
  
Siddhesh Poyarekar Jan. 16, 2019, 3:35 a.m. UTC | #6
On 16/01/19 3:06 AM, Richard Stallman wrote:

> I think so.  It does not belong in any other GNU package as far as I
> know.  I am sure maintainers of various GNU packages will want to use
> it.

OK.  Does anyone else have thoughts to the contrary?  If not I'll create 
a separate project with this.  In that case the only thing we need to 
agree on is the transition plan for glibc.


Siddhesh
  
Rical Jasan Jan. 16, 2019, 5:27 a.m. UTC | #7
On 01/15/2019 07:35 PM, Siddhesh Poyarekar wrote:
> On 16/01/19 3:06 AM, Richard Stallman wrote:
>> I think so.  It does not belong in any other GNU package as far as I
>> know.  I am sure maintainers of various GNU packages will want to use
>> it.
> OK.  Does anyone else have thoughts to the contrary?  If not I'll create
> a separate project with this.  In that case the only thing we need to
> agree on is the transition plan for glibc.

I think this is more well-suited to being a separate package.  It may be
by-glibc-for-glibc right now, but I see it as having value for other
projects too.  It certainly seems like something we would want to share,
and there is a lot of potential behind the idea.

I've taken a look at the script but haven't played with it, and Python
isn't really my strong suit.  As much as I'd like to try and help hack
on it, my spare development time is practically nil right now, but I
wanted to at least lend a voice.

Rical
  
Siddhesh Poyarekar Jan. 16, 2019, 3:28 p.m. UTC | #8
> I think this is more well-suited to being a separate package.  It may be
> by-glibc-for-glibc right now, but I see it as having value for other
> projects too.  It certainly seems like something we would want to share,
> and there is a lot of potential behind the idea.
>
> I've taken a look at the script but haven't played with it, and Python
> isn't really my strong suit.  As much as I'd like to try and help hack
> on it, my spare development time is practically nil right now, but I
> wanted to at least lend a voice.


Thank you for sharing your opinion.  I have now requested a project
space on savannah under the name vcs2changelog.  Hopefully it will be
approved soon so that I can start breaking the script down into a proper
backend+frontend in multiple files.


Siddhesh
  
Siddhesh Poyarekar Jan. 16, 2019, 3:47 p.m. UTC | #9
Hmm, and now the GNU mailserver seems to have decided it doesn't like my
email addresses and is bouncing my emails:

Remote-MTA: dns; eggs.gnu.org
Diagnostic-Code: smtp; 550 Previous (cached) callout verification failure

Siddhesh


On 16/01/19 8:58 PM, Siddhesh Poyarekar wrote:
> 
>> I think this is more well-suited to being a separate package.  It may be
>> by-glibc-for-glibc right now, but I see it as having value for other
>> projects too.  It certainly seems like something we would want to share,
>> and there is a lot of potential behind the idea.
>>
>> I've taken a look at the script but haven't played with it, and Python
>> isn't really my strong suit.  As much as I'd like to try and help hack
>> on it, my spare development time is practically nil right now, but I
>> wanted to at least lend a voice.
> 
> 
> Thank you for sharing your opinion.  I have now requested a project
> space on savannah under the name vcs2changelog.  Hopefully it will be
> approved soon so that I can start breaking the script down into a proper
> backend+frontend in multiple files.
> 
> 
> Siddhesh
> 
>
  
Alfred M. Szmidt Jan. 16, 2019, 4:54 p.m. UTC | #10
> I think so.  It does not belong in any other GNU package as far as I
   > know.  I am sure maintainers of various GNU packages will want to use
   > it.

   OK.  Does anyone else have thoughts to the contrary?  If not I'll create 
   a separate project with this.  In that case the only thing we need to 
   agree on is the transition plan for glibc.

gnulib?
  
Siddhesh Poyarekar Jan. 16, 2019, 4:59 p.m. UTC | #11
On 16/01/19 10:24 PM, Alfred M. Szmidt wrote:
>    > I think so.  It does not belong in any other GNU package as far as I
>    > know.  I am sure maintainers of various GNU packages will want to use
>    > it.
>
>    OK.  Does anyone else have thoughts to the contrary?  If not I'll create 
>    a separate project with this.  In that case the only thing we need to 
>    agree on is the transition plan for glibc.
>
> gnulib?

The idea I have in mind now probably goes beyond just build-aux in
gnulib.  It ought to have support for multiple backends (at least svn
and git) and support multiple languages, so it might end up blowing up
beyond gnulib's intent.


That is unless you're suggesting that it would be useful to have a
separate vcs2changelog directory inside gnulib that does all of this. 
That doesn't seem too bad an idea if the gnulib maintainers are OK with
it.  Saves me the trouble of setting up a new repo too...


Siddhesh
  
Paul Eggert Jan. 16, 2019, 5:21 p.m. UTC | #12
On 1/16/19 8:59 AM, Siddhesh Poyarekar wrote:
> That is unless you're suggesting that it would be useful to have a
> separate vcs2changelog directory inside gnulib that does all of this.
> That doesn't seem too bad an idea if the gnulib maintainers are OK with
> it.  Saves me the trouble of setting up a new repo too...

I'd rather that there was one script for this rather than two. Currently 
the Gnulib script is used by many other projects (GNU Emacs, coreutils, 
...) and we should be looking to use a new-and-improved script in those 
other projects too. So I'd welcome the idea of putting this in Gnulib.
  
Joseph Myers Jan. 16, 2019, 5:29 p.m. UTC | #13
On Wed, 16 Jan 2019, Paul Eggert wrote:

> I'd rather that there was one script for this rather than two. Currently the
> Gnulib script is used by many other projects (GNU Emacs, coreutils, ...) and
> we should be looking to use a new-and-improved script in those other projects
> too. So I'd welcome the idea of putting this in Gnulib.

Note that the scripts do different things.  The gnulib script is for the 
case where the commit messages include ChangeLog-like content.  The glibc 
one is for the case where the commit messages do not try to list changed 
entities or what changed in those entities, but where an 
automatically-generated list of changed entities is considered good enough 
if the listing of such entities is accurate enough as applied to commits 
in the particular GNU project in question.  (But I'd hope most of the core 
code in the glibc script could be shared by other projects, even if there 
are some bits specific to glibc-specific source code constructs.  Cf. how 
we use the gnulib update-copyright script, with just a local wrapper that 
arranges to call it with the right options for each glibc source file.)
  
Paul Eggert Jan. 16, 2019, 5:44 p.m. UTC | #14
On 1/16/19 9:29 AM, Joseph Myers wrote:
> Note that the scripts do different things.  The gnulib script is for the
> case where the commit messages include ChangeLog-like content.  The glibc
> one is for the case where the commit messages do not try to list changed
> entities or what changed in those entities, but where an
> automatically-generated list of changed entities is considered good enough

Yes, the idea is that Emacs, coreutils, etc. will switch to the Glibc 
approach. Nobody prefers the current Gnulib approach, as far as I know; 
it's used only because the Gnulib script was easier to write.
  
Siddhesh Poyarekar Jan. 16, 2019, 5:56 p.m. UTC | #15
On 16/01/19 10:51 PM, Paul Eggert wrote:
> On 1/16/19 8:59 AM, Siddhesh Poyarekar wrote:
>> That is unless you're suggesting that it would be useful to have a
>> separate vcs2changelog directory inside gnulib that does all of this.
>> That doesn't seem too bad an idea if the gnulib maintainers are OK with
>> it.  Saves me the trouble of setting up a new repo too...
> 
> I'd rather that there was one script for this rather than two. Currently
> the Gnulib script is used by many other projects (GNU Emacs, coreutils,
> ...) and we should be looking to use a new-and-improved script in those
> other projects too. So I'd welcome the idea of putting this in Gnulib.

I agree with that idea, but the single script will most likely blow up
into a set of scripts with the driver and then parsers for backends (c,
c++, etc.) and quirks (e.g. macros that break the parser) for specific
projects.  Would that be acceptable?  If yes then I'll post this patch
to the gnulib list inside build-aux/vcs2changelog/* and also put the
current git-to-changelog into my scripts as a compatibility mode.

Siddhesh
  
Paul Eggert Jan. 16, 2019, 6:02 p.m. UTC | #16
On 1/16/19 9:56 AM, Siddhesh Poyarekar wrote:
> I agree with that idea, but the single script will most likely blow up
> into a set of scripts with the driver and then parsers for backends (c,
> c++, etc.) and quirks (e.g. macros that break the parser) for specific
> projects.  Would that be acceptable?

Sounds good to me.
  

Patch

diff --git a/scripts/gen-changed-entities.py b/scripts/gen-changed-entities.py
new file mode 100755
index 0000000000..6d2031ddb7
--- /dev/null
+++ b/scripts/gen-changed-entities.py
@@ -0,0 +1,1099 @@ 
+#!/usr/bin/python3
+# Copyright (C) 2019 Free Software Foundation, Inc.
+# This file is part of the GNU C Library.
+#
+# The GNU C Library is free software; you can redistribute it and/or
+# modify it under the terms of the GNU Lesser General Public
+# License as published by the Free Software Foundation; either
+# version 2.1 of the License, or (at your option) any later version.
+#
+# The GNU C Library is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# Lesser General Public License for more details.
+#
+# You should have received a copy of the GNU Lesser General Public
+# License along with the GNU C Library; if not, see
+# <http://www.gnu.org/licenses/>.
+''' Generate a ChangeLog style output based on the git log.
+
+This script takes two revisions as input and generates a ChangeLog style output
+for all revisions between the two revisions.  This output is intended to be an
+approximation and not the exact ChangeLog.
+
+At a high level, the script enumerates all C source files (*.c and *.h) and
+builds a tree of top level objects within macro conditionals.  The top level
+objects the script currently attempts to identify are:
+
+    - Include statements
+    - Macro definitions and undefs
+    - Declarations and definitions of variables and functions
+    - Composite types
+
+The script attempts to identify quirks typically used in glibc sources such as
+the symbol hack macro calls that don't use a semicolon and tries to adjust for
+them.
+
+Known Limitations:
+
+    - Does not identify changes in or to comments.  Comments are simply stripped
+      out.
+    - Weird nesting of macro conditionals may break things.  Attempts have been
+      made to try and maintain state across macro conditional scopes, but
+      there's still scope to fool the script.
+    - Does not identify changes within functions.
+'''
+import subprocess
+import sys
+import os
+import re
+from enum import Enum
+
+# General Utility functions.
+def eprint(*args, **kwargs):
+    ''' Print to stderr.
+    '''
+    print(*args, file=sys.stderr, **kwargs)
+
+
+debug = False
+def debug_print(*args, **kwargs):
+    ''' Convenience function to print diagnostic information in the program.
+    '''
+    if debug:
+        eprint(*args, **kwargs)
+
+
+def usage(name):
+    ''' Print program usage.
+    '''
+    eprint("usage: %s <from-ref> <to-ref>" % name)
+    sys.exit(os.EX_USAGE)
+
+
+def decode(string):
+    ''' Attempt to decode a string.
+
+    Decode a string read from the source file.  The multiple attempts are needed
+    due to the presence of the page break characters and some tests in locales.
+    '''
+    codecs = ['utf8', 'latin1', 'cp1252']
+
+    for i in codecs:
+        try:
+            return string.decode(i)
+        except UnicodeDecodeError:
+            pass
+
+    eprint('Failed to decode: %s' % string)
+
+
+#------------------------------------------------------------------------------
+# C Parser.
+#------------------------------------------------------------------------------
+
+
+def new_block(name, type, contents, parent, flags = 0):
+    '''  Create a new code block with the parent as PARENT.
+
+    The code block is a basic structure around which the tree representation of
+    the source code is built.  It has the following attributes:
+
+    - name: A name to refer it by in the ChangeLog
+    - type: Any one of the following types in BLOCK_TYPE.
+    - contents: The contents of the block.  For a block of types file or
+      macro_cond, this would be a list of blocks that it nests.  For other types
+      it is a list with a single string specifying its contents.
+    - parent: This is the parent of the current block, useful in setting up
+      #elif or #else blocks in the tree.
+    - flags: A special field to indicate some properties of the block. See
+      BLOCK_FLAGS for values.
+    '''
+    block = {}
+    block['matched'] = False
+    block['name'] = name
+    block['type'] = type
+    block['contents'] = contents
+    block['parent'] = parent
+    if parent:
+        parent['contents'].append(block)
+
+    block['flags'] = flags
+
+    return block
+
+
+class block_flags(Enum):
+    ''' Flags for the code block.
+    '''
+    else_block = 1
+    macro_defined = 2
+    macro_redefined = 3
+
+
+class block_type(Enum):
+    ''' Type of code block.
+    '''
+    file = 1
+    macro_cond = 2
+    macro_def = 3
+    macro_undef = 4
+    macro_include = 5
+    macro_info = 6
+    decl = 7
+    func = 8
+    composite = 9
+    macrocall = 10
+    fndecl = 11
+    assign = 12
+    struct = 13
+    union = 14
+    enum = 15
+
+
+# A dictionary describing what each action (add, modify, delete) show up as in
+# the ChangeLog output.
+actions = {0:{'new': 'New', 'mod': 'Modified', 'del': 'Remove'},
+           block_type.file:{'new': 'New file', 'mod': 'Modified file',
+                            'del': 'Remove file'},
+           block_type.macro_cond:{'new': 'New', 'mod': 'Modified',
+                                  'del': 'Remove'},
+           block_type.macro_def:{'new': 'New', 'mod': 'Modified',
+                                 'del': 'Remove'},
+           block_type.macro_include:{'new': 'Include file', 'mod': 'Modified',
+                                     'del': 'Remove include'},
+           block_type.macro_info:{'new': 'New preprocessor message',
+                                  'mod': 'Modified', 'del': 'Remove'},
+           block_type.decl:{'new': 'New', 'mod': 'Modified', 'del': 'Remove'},
+           block_type.func:{'new': 'New function', 'mod': 'Modified function',
+                 'del': 'Remove function'},
+           block_type.composite:{'new': 'New', 'mod': 'Modified',
+                                 'del': 'Remove'},
+           block_type.struct:{'new': 'New struct', 'mod': 'Modified struct',
+                                 'del': 'Remove struct'},
+           block_type.union:{'new': 'New union', 'mod': 'Modified union',
+                                 'del': 'Remove union'},
+           block_type.enum:{'new': 'New enum', 'mod': 'Modified enum',
+                                 'del': 'Remove enum'},
+           block_type.macrocall:{'new': 'New', 'mod': 'Modified',
+                                 'del': 'Remove'},
+           block_type.fndecl:{'new': 'New function', 'mod': 'Modified',
+                              'del': 'Remove'},
+           block_type.assign:{'new': 'New', 'mod': 'Modified', 'del': 'Remove'}}
+
+# Regular expressions.
+
+# The __attribute__ are written in a bunch of different ways in glibc.
+ATTRIBUTE = \
+r'(((attribute_\w+)|(__attribute__\s*\(\([^;]+\)\))|(weak_function)|(_*ATTRIBUTE_*\w+)|(asm\s*\([^\)]+\)))\s*)*'
+
+# Different Types that we recognize.  This is specifically to filter out any
+# macro hack types.
+TYPE_RE = r'(\w+|(ElfW\(\w+\)))'
+
+# Function regex
+FUNC_RE = re.compile(ATTRIBUTE + r'\s*(\w+)\s*\([^(][^{]+\)\s*{')
+
+# The macrocall_re peeks into the next line to ensure that it doesn't eat up
+# a FUNC by accident.  The func_re regex is also quite crude and only
+# intends to ensure that the function name gets picked up correctly.
+MACROCALL_RE = re.compile(r'(\w+)\s*(\(.*\))*$')
+
+# Composite types such as structs and unions.
+COMPOSITE_RE = re.compile(r'(struct|union|enum)\s*(\w*)\s*{')
+
+# Static assignments.
+ASSIGN_RE = re.compile(r'(\w+)\s*(\[[^\]]*\])*\s*([^\s]*attribute[\s\w()]+)?\s*=')
+
+# Function Declarations.
+FNDECL_RE = re.compile(r'\s*(\w+)\s*\([^\(][^;]*\)\s*(__THROW)?\s*(asm\s*\([?)]\))?\s*' + ATTRIBUTE + ';')
+
+# Function pointer typedefs.
+TYPEDEF_FN_RE = re.compile(r'\(\*(\w+)\)\s*\([^)]+\);')
+
+# Simple decls.
+DECL_RE = re.compile(r'(\w+)(\[\w*\])*\s*' + ATTRIBUTE + ';')
+
+# __typeof decls.
+TYPEOF_DECL_RE = re.compile(r'__typeof\s*\([\w\s]+\)\s*(\w+)\s*' + ATTRIBUTE + ';')
+
+
+def remove_extern_c(op):
+    ''' Process extern "C"/"C++" block nesting.
+
+    The extern "C" nesting does not add much value so it's safe to almost always
+    drop it.  Also drop extern "C++"
+    '''
+    new_op = []
+    nesting = 0
+    extern_nesting = 0
+    for l in op:
+        if '{' in l:
+            nesting = nesting + 1
+        if re.match(r'extern\s*"C"\s*{', l):
+            extern_nesting = nesting
+            continue
+        if '}' in l:
+            nesting = nesting - 1
+            if nesting < extern_nesting:
+                extern_nesting = 0
+                continue
+        new_op.append(l)
+
+    # Now drop all extern C++ blocks.
+    op = new_op
+    new_op = []
+    nesting = 0
+    extern_nesting = 0
+    in_cpp = False
+    for l in op:
+        if re.match(r'extern\s*"C\+\+"\s*{', l):
+            nesting = nesting + 1
+            in_cpp = True
+
+        if in_cpp:
+            if '{' in l:
+                nesting = nesting + 1
+            if '}' in l:
+                nesting = nesting - 1
+        if nesting == 0:
+            new_op.append(l)
+
+    return new_op
+
+
+def remove_comments(op):
+    ''' Remove comments.
+
+    Return OP by removing all comments from it.
+    '''
+    debug_print('REMOVE COMMENTS')
+
+    sep='\n'
+    opstr = sep.join(op)
+    opstr = re.sub(r'/\*.*?\*/', r'', opstr, flags=re.MULTILINE | re.DOTALL)
+    opstr = re.sub(r'\\\n', r' ', opstr, flags=re.MULTILINE | re.DOTALL)
+    new_op = list(filter(None, opstr.split(sep)))
+
+    return new_op
+
+
+def normalize_condition(name):
+    ''' Make some minor transformations on macro conditions to make them more
+    readable.
+    '''
+    # Negation with a redundant bracket.
+    name = re.sub(r'!\s*\(\s*(\w+)\s*\)', r'! \1', name)
+    # Pull in negation of equality.
+    name = re.sub(r'!\s*\(\s*(\w+)\s*==\s*(\w+)\)', r'\1 != \2', name)
+    # Pull in negation of inequality.
+    name = re.sub(r'!\s*\(\s*(\w+)\s*!=\s*(\w+)\)', r'\1 == \2', name)
+    # Fix simple double negation.
+    name = re.sub(r'!\s*\(\s*!\s*(\w+)\s*\)', r'\1', name)
+    # Similar, but nesting a complex expression.  Because of the greedy match,
+    # this matches only the outermost brackets.
+    name = re.sub(r'!\s*\(\s*!\s*\((.*)\)\s*\)$', r'\1', name)
+    return name
+
+
+def parse_preprocessor(op, loc, code, start = ''):
+    ''' Parse a preprocessor directive.
+
+    In case a preprocessor condition (i.e. if/elif/else), create a new code
+    block to nest code into and in other cases, identify and add entities suchas
+    include files, defines, etc.
+
+    - OP is the string array for the file
+    - LOC is the first unread location in CUR
+    - CODE is the block to which we add this function
+    - START is the string that should continue to be expanded in case we step
+      into a new macro scope.
+
+    - Returns: The next location to be read in the array.
+    '''
+    cur = op[loc]
+    loc = loc + 1
+    endblock = False
+
+    debug_print('PARSE_MACRO: %s' % cur)
+
+    # Remove the # and strip spaces again.
+    cur = cur[1:].strip()
+
+    # Include file.
+    if cur.find('include') == 0:
+        m = re.search(r'include\s*["<]?([^">]+)[">]?', cur)
+        new_block(m.group(1), block_type.macro_include, [cur], code)
+
+    # Macro definition.
+    if cur.find('define') == 0:
+        m = re.search(r'define\s+([a-zA-Z0-9_]+)', cur)
+        name = m.group(1)
+        exists = False
+        # Find out if this is a redefinition.
+        for c in code['contents']:
+            if c['name'] == name and c['type'] == block_type.macro_def:
+                c['flags'] = block_flags.macro_redefined
+                exists = True
+                break
+        if not exists:
+            new_block(m.group(1), block_type.macro_def, [cur], code,
+                      block_flags.macro_defined)
+            # Add macros as we encounter them.
+            SYM_MACROS.append(m.group(1))
+
+    # Macro undef.
+    if cur.find('undef') == 0:
+        m = re.search(r'undef\s+([a-zA-Z0-9_]+)', cur)
+        new_block(m.group(1), block_type.macro_def, [cur], code)
+
+    # #error and #warning macros.
+    if cur.find('error') == 0 or cur.find('warning') == 0:
+        m = re.search(r'(error|warning)\s+"?(.*)"?', cur)
+        if m:
+            name = m.group(2)
+        else:
+            name = '<blank>'
+        new_block(name, block_type.macro_info, [cur], code)
+
+    # Start of an #if or #ifdef block.
+    elif cur.find('if') == 0:
+        rem = re.sub(r'ifndef', r'!', cur).strip()
+        rem = re.sub(r'(ifdef|defined|if)', r'', rem).strip()
+        rem = normalize_condition(rem)
+        ifdef = new_block(rem, block_type.macro_cond, [], code)
+        ifdef['headcond'] = ifdef
+        ifdef['start'] = start
+        loc = c_parse(op, loc, ifdef, start)
+
+    # End the previous #if/#elif and begin a new block.
+    elif cur.find('elif') == 0 and code['parent']:
+        rem = normalize_condition(re.sub(r'(elif|defined)', r'', cur).strip())
+        # The #else and #elif blocks should go into the current block's parent.
+        ifdef = new_block(rem, block_type.macro_cond, [], code['parent'])
+        ifdef['headcond'] = code['headcond']
+        loc = c_parse(op, loc, ifdef, code['headcond']['start'])
+        endblock = True
+
+    # End the previous #if/#elif and begin a new block.
+    elif cur.find('else') == 0 and code['parent']:
+        name = normalize_condition('!(' + code['name'] + ')')
+        ifdef = new_block(name, block_type.macro_cond, [], code['parent'],
+                          block_flags.else_block)
+        ifdef['headcond'] = code['headcond']
+        loc = c_parse(op, loc, ifdef, code['headcond']['start'])
+        endblock = True
+
+    elif cur.find('endif') == 0 and code['parent']:
+        # Insert an empty else block if there isn't one.
+        if code['flags'] != block_flags.else_block:
+            name = normalize_condition('!(' + code['name'] + ')')
+            ifdef = new_block(name, block_type.macro_cond, [], code['parent'],
+                              block_flags.else_block)
+            ifdef['headcond'] = code['headcond']
+            loc = c_parse(op, loc - 1, ifdef, code['headcond']['start'])
+        endblock = True
+
+    return (loc, endblock)
+
+
+def fast_forward_scope(cur, op, loc):
+    ''' Consume lines in a code block.
+
+    Consume all lines of a block of code such as a composite type declaration or
+    a function declaration.
+
+    - CUR is the string to consume this expression from
+    - OP is the string array for the file
+    - LOC is the first unread location in CUR
+
+    - Returns: The next location to be read in the array as well as the updated
+      value of CUR, which will now have the body of the function or composite
+      type.
+    '''
+    nesting = cur.count('{') - cur.count('}')
+    while nesting > 0 and loc < len(op):
+        cur = cur + ' ' + op[loc]
+
+        nesting = nesting + op[loc].count('{')
+        nesting = nesting - op[loc].count('}')
+        loc = loc + 1
+
+    return (cur, loc)
+
+
+def parse_decl(name, cur, op, loc, code, blocktype, match):
+    ''' Parse a top level declaration.
+
+    All types of declarations except function declarations.
+
+    - NAME is the name of the declarated entity
+    - CUR is the string to consume this expression from
+    - OP is the string array for the file
+    - LOC is the first unread location in CUR
+    - CODE is the block to which we add this function
+    - BLOCKTYPE is the type of the block
+    - MATCH is the regex match object.
+
+    - Returns: The next location to be read in the array.
+    '''
+    debug_print('FOUND DECL: %s' % name)
+    new_block(name, blocktype, [cur], code)
+
+    return loc
+
+
+def parse_assign(name, cur, op, loc, code, blocktype, match):
+    ''' Parse an assignment statement.
+
+    This includes array assignments.
+
+    - NAME is the name of the assigned entity
+    - CUR is the string to consume this expression from
+    - OP is the string array for the file
+    - LOC is the first unread location in CUR
+    - CODE is the block to which we add this
+    - BLOCKTYPE is the type of the block
+    - MATCH is the regex match object.
+
+    - Returns: The next location to be read in the array.
+    '''
+    debug_print('FOUND ASSIGN: %s' % name)
+    # Lap up everything up to semicolon.
+    while ';' not in cur and loc < len(op):
+        cur = op[loc]
+        loc = loc + 1
+
+    new_block(name, blocktype, [cur], code)
+
+    return loc
+
+
+def parse_composite(name, cur, op, loc, code, blocktype, match):
+    ''' Parse a composite type.
+
+    Match declaration of a composite type such as a sruct or a union..
+
+    - NAME is the name of the composite type
+    - CUR is the string to consume this expression from
+    - OP is the string array for the file
+    - LOC is the first unread location in CUR
+    - CODE is the block to which we add this
+    - BLOCKTYPE is the type of the block
+    - MATCH is the regex match object.
+
+    - Returns: The next location to be read in the array.
+    '''
+    # Lap up all of the struct definition.
+    (cur, loc) = fast_forward_scope(cur, op, loc)
+
+    if not name:
+        if 'typedef' in cur:
+            name = re.sub(r'.*}\s*(\w+);$', r'\1', cur)
+        else:
+            name= '<anoymous>'
+
+    ctype = match.group(1)
+
+    if ctype == 'struct':
+        blocktype = block_type.struct
+    if ctype == 'enum':
+        blocktype = block_type.enum
+    if ctype == 'union':
+        blocktype = block_type.union
+
+    new_block(name, blocktype, [cur], code)
+
+    return loc
+
+
+def parse_func(name, cur, op, loc, code, blocktype, match):
+    ''' Parse a function.
+
+    Match a function definition.
+
+    - NAME is the name of the function
+    - CUR is the string to consume this expression from
+    - OP is the string array for the file
+    - LOC is the first unread location in CUR
+    - CODE is the block to which we add this
+    - BLOCKTYPE is the type of the block
+    - MATCH is the regex match object.
+
+    - Returns: The next location to be read in the array.
+    '''
+    debug_print('FOUND FUNC: %s' % name)
+
+    # Consume everything up to the ending brace of the function.
+    (cur, loc) = fast_forward_scope(cur, op, loc)
+
+    new_block(name, blocktype, [cur], code)
+
+    return loc
+
+
+SYM_MACROS = []
+def parse_macrocall(cur, op, loc, code, blocktype, match):
+    ''' Parse a macro call.
+
+    Match a symbol hack macro calls that get added without semicolons.
+
+    - CUR is the string to consume this expression from
+    - OP is the string array for the file
+    - LOC is the first unread location in CUR
+    - CODE is the block to which we add this
+    - BLOCKTYPE is the type of the block
+    - MATCH is the regex match object.
+
+    - Returns: The next location to be read in the array.
+    '''
+
+    # First we have the macros for symbol hacks and all macros we identified so
+    # far.
+    if cur.count('(') != cur.count(')'):
+        return cur, loc
+    if loc < len(op) and '{' in op[loc]:
+        return cur, loc
+
+    found = re.search(MACROCALL_RE, cur)
+    if found:
+        sym = found.group(1)
+        name = found.group(2)
+        if sym in SYM_MACROS:
+            debug_print('FOUND MACROCALL: %s (%s)' % (sym, name))
+            new_block(sym, blocktype, [cur], code)
+            return '', loc
+
+    # Next, there could be macros that get called right inside their #ifdef, but
+    # without the semi-colon.
+    if cur.strip() == code['name'].strip():
+        debug_print('FOUND MACROCALL (without brackets): %s' % (cur))
+        new_block(cur, blocktype, [cur], code)
+        return '',loc
+
+    return cur, loc
+
+
+# Regular expressions and token consumption functions for expressions we
+# encounter in C.
+c_expr_parsers = [
+        {'regex' : COMPOSITE_RE, 'func' : parse_composite, 'name' : 2,
+            'type' : block_type.composite},
+        {'regex' : ASSIGN_RE, 'func' : parse_assign, 'name' : 1,
+            'type' : block_type.assign},
+        {'regex' : TYPEOF_DECL_RE, 'func' : parse_decl, 'name' : 1,
+            'type' : block_type.decl},
+        {'regex' : TYPEDEF_FN_RE, 'func' : parse_decl, 'name' : 1,
+            'type' : block_type.decl},
+        {'regex' : FNDECL_RE, 'func' : parse_decl, 'name' : 1,
+            'type' : block_type.fndecl},
+        {'regex' : FUNC_RE, 'func' : parse_func, 'name' : 8,
+            'type' : block_type.func},
+        {'regex' : DECL_RE, 'func' : parse_decl, 'name' : 1,
+            'type' : block_type.decl},
+        # Special case at the end: macro call.
+        {'regex' : None, 'func' : parse_macrocall, 'name' : 1,
+            'type' : block_type.macrocall}]
+
+
+def parse_c_expr(cur, op, loc, code):
+    ''' Parse a C expression.
+
+    CUR is the string to be parsed, which continues to grow until a match is
+    found.  OP is the string array and LOC is the first unread location in the
+    string array.  CODE is the block in which any identified expressions should
+    be added.
+    '''
+    debug_print('PARSING: %s' % cur)
+
+    for p in c_expr_parsers:
+        if not p['regex']:
+            return p['func'](cur, op, loc, code, p['type'], found)
+
+        found = re.search(p['regex'], cur)
+        if found:
+            return '', p['func'](found.group(p['name']), cur, op, loc, code,
+                                    p['type'], found)
+
+    return cur, loc
+
+
+# These are macros that break parsing and don't have any intuitie way to get
+# around.  Just replace them with something parseable.
+PROBLEM_MACROS = \
+    [{'orig': r'ElfW\((\w+)\)', 'sub': r'\1__ELF_NATIVE_CLASS_t'},
+     {'orig': r'(libc_freeres_fn)\s*\((\w+)\)', 'sub': r'static void \1__\2 (void)'},
+     {'orig': r'(IMPL)\s*\((\w+), .*\)$', 'sub': r'static void \1__\2 (void) {}'},
+     {'orig': r'__(BEGIN|END)_DECLS', 'sub': r''}]
+
+
+def expand_problematic_macros(cur):
+    ''' Replace problem macros with their substitutes in CUR.
+    '''
+    for p in PROBLEM_MACROS:
+        cur = re.sub(p['orig'], p['sub'], cur)
+
+    return cur
+
+
+def c_parse(op, loc, code, start = ''):
+    '''
+    Parse the file line by line.  The function assumes a mostly GNU coding
+    standard compliant input so it might barf with anything that is eligible for
+    the Obfuscated C code contest.
+
+    The basic idea of the parser is to identify macro conditional scopes and
+    definitions, includes, etc. and then parse the remaining C code in the
+    context of those macro scopes.  The parser does not try to understand the
+    semantics of the code or even validate its syntax.  It only records high
+    level symbols in the source and makes a tree structure to indicate the
+    declaration/definition of those symbols and their scope in the macro
+    definitions.
+
+    OP is the string array.
+    LOC is the first unparsed line.
+    CODE is the block scope within which the parsing is currently going on.
+    START is the string with which this parsing should start.
+    '''
+    cur = start
+    endblock = False
+    saved_cur = ''
+    saved_loc = 0
+    endblock_loc = loc
+
+    while loc < len(op):
+        nextline = op[loc]
+
+        # Macros.
+        if nextline[0] == '#':
+            (loc, endblock) = parse_preprocessor(op, loc, code, cur)
+            if endblock:
+                endblock_loc = loc
+        # Rest of C Code.
+        else:
+            cur = cur + ' ' + nextline
+            cur = expand_problematic_macros(cur).strip()
+            cur, loc = parse_c_expr(cur, op, loc + 1, code)
+
+        if endblock and not cur:
+            # If we are returning from the first #if block, we want to proceed
+            # beyond the current block, not repeat it for any preceding blocks.
+            if code['headcond'] == code:
+                return loc
+            else:
+                return endblock_loc
+
+    return loc
+
+def drop_empty_blocks(tree):
+    ''' Drop empty macro conditional blocks.
+    '''
+    newcontents = []
+
+    for x in tree['contents']:
+        if x['type'] != block_type.macro_cond or len(x['contents']) > 0:
+            newcontents.append(x)
+
+    for t in newcontents:
+        if t['type'] == block_type.macro_cond:
+            drop_empty_blocks(t)
+
+    tree['contents'] = newcontents
+
+
+def consolidate_tree_blocks(tree):
+    ''' Consolidate common macro conditional blocks.
+
+    Get macro conditional blocks at the same level but scatterred across the
+    file together into a single common block to allow for better comparison.
+    '''
+    # Nothing to do for non-nesting blocks.
+    if tree['type'] != block_type.macro_cond \
+            and tree['type'] != block_type.file:
+        return
+
+    # Now for nesting blocks, get the list of unique condition names and
+    # consolidate code under them.  The result also bunches up all the
+    # conditions at the top.
+    newcontents = []
+
+    macros = [x for x in tree['contents'] \
+              if x['type'] == block_type.macro_cond]
+    macro_names = sorted(set([x['name'] for x in macros]))
+    for m in macro_names:
+        nc = [x['contents'] for x in tree['contents'] if x['name'] == m \
+                and x['type'] == block_type.macro_cond]
+        b = new_block(m, block_type.macro_cond, sum(nc, []), tree)
+        consolidate_tree_blocks(b)
+        newcontents.append(b)
+
+    newcontents.extend([x for x in tree['contents'] \
+                        if x['type'] != block_type.macro_cond])
+
+    tree['contents'] = newcontents
+
+
+def compact_tree(tree):
+    ''' Try to reduce the tree to its minimal form.
+
+    A source code tree in its simplest form may have a lot of duplicated
+    information that may be difficult to compare and come up with a minimal
+    difference.
+    '''
+
+    # First, drop all empty blocks.
+    drop_empty_blocks(tree)
+
+    # Macro conditions that nest the entire file aren't very interesting.  This
+    # should take care of the header guards.
+    if tree['type'] == block_type.file \
+            and len(tree['contents']) == 1 \
+            and tree['contents'][0]['type'] == block_type.macro_cond:
+        tree['contents'] = tree['contents'][0]['contents']
+
+    # Finally consolidate all macro conditional blocks.
+    consolidate_tree_blocks(tree)
+
+
+def build_macrocalls():
+    ''' Build a list of macro calls used for symbol versioning and attributes.
+
+    glibc uses a set of macro calls that do not end with a semi-colon and hence
+    breaks our parser.  Identify those calls from include/libc-symbols.h and
+    filter them out.
+    '''
+    with open('../include/libc-symbols.h') as macrofile:
+        global SYM_MACROS
+        op = macrofile.readlines()
+        op = remove_comments(op)
+        SYM_MACROS = [re.sub(r'.*define (\w+).*', r'\1', x[:-1]) for x in op \
+                      if 'define ' in x]
+
+
+def c_parse_output(op):
+    ''' File parser.
+
+    Parse the input array of lines OP and generate a tree structure to
+    represent the file.  This tree structure is then used for comparison between
+    the old and new file.
+    '''
+    build_macrocalls()
+    tree = new_block('', block_type.file, [], None)
+    op = remove_comments(op)
+    op = remove_extern_c(op)
+    op = [re.sub(r'#\s+', '#', x) for x in op]
+    op = c_parse(op, 0, tree)
+    compact_tree(tree)
+    c_dump_tree(tree, 0)
+
+    return tree
+
+
+def print_change(tree, action, prologue = ''):
+    ''' Print the nature of the differences found in the tree compared to the
+    other tree.  TREE is the tree that changed, action is what the change was
+    (Added, Removed, Modified) and prologue specifies the macro scope the change
+    is in.  The function calls itself recursively for all macro condition tree
+    nodes.
+    '''
+
+    if tree['type'] != block_type.macro_cond:
+        print('\t%s(%s): %s.' % (prologue, tree['name'], action))
+        return
+
+    prologue = '%s[%s]' % (prologue, tree['name'])
+    for t in tree['contents']:
+        if t['type'] == block_type.macro_cond:
+            print_change(t, action, prologue)
+        else:
+            print('\t%s(%s): %s.' % (prologue, t['name'], action))
+
+
+def c_compare_trees(left, right, prologue = ''):
+    ''' Compare two trees and print the difference.
+
+    This routine is the entry point to compare two trees and print out their
+    differences.  LEFT and RIGHT will always have the same name and type,
+    starting with block_type.file and '' at the top level.
+    '''
+
+    if left['type'] == block_type.macro_cond or left['type'] == block_type.file:
+
+        if left['type'] == block_type.macro_cond:
+            prologue = '%s[%s]' % (prologue, left['name'])
+
+        # Make sure that everything in the left tree exists in the right tree.
+        for cl in left['contents']:
+            found = False
+            for cr in right['contents']:
+                if not cl['matched'] and not cr['matched'] and \
+                        cl['name'] == cr['name'] and cl['type'] == cr['type']:
+                    cl['matched'] = cr['matched'] = True
+                    c_compare_trees(cl, cr, prologue)
+                    found = True
+                    break
+            if not found:
+                print_change(cl, actions[cl['type']]['del'], prologue)
+
+        # ... and vice versa.  This time we only need to look at unmatched
+        # contents.
+        for cr in right['contents']:
+            if not cr['matched']:
+                print_change(cr, actions[cr['type']]['new'], prologue)
+    else:
+        if left['contents'] != right['contents']:
+            print_change(left, actions[left['type']]['mod'], prologue)
+
+
+def c_dump_tree(tree, indent):
+    ''' Print the entire tree.
+    '''
+    if not debug:
+        return
+
+    if tree['type'] == block_type.macro_cond or tree['type'] == block_type.file:
+        print('%sScope: %s' % (' ' * indent, tree['name']))
+        for c in tree['contents']:
+            c_dump_tree(c, indent + 4)
+        print('%sEndScope: %s' % (' ' * indent, tree['name']))
+    else:
+        if tree['type'] == block_type.func:
+            print('%sFUNC: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.composite:
+            print('%sCOMPOSITE: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.assign:
+            print('%sASSIGN: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.fndecl:
+            print('%sFNDECL: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.decl:
+            print('%sDECL: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.macrocall:
+            print('%sMACROCALL: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.macro_def:
+            print('%sDEFINE: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.macro_include:
+            print('%sINCLUDE: %s' % (' ' * indent, tree['name']))
+        elif tree['type'] == block_type.macro_undef:
+            print('%sUNDEF: %s' % (' ' * indent, tree['name']))
+        else:
+            print('%sMACRO LEAF: %s' % (' ' * indent, tree['name']))
+
+
+def c_compare(oldfile, newfile):
+    ''' Entry point for the C backend.
+
+    Parse the two files into trees and compare them.  Print the result of the
+    comparison in the ChangeLog-like format.
+    '''
+    debug_print('LEFT TREE')
+    debug_print('-' * 80)
+    left = c_parse_output(oldfile)
+
+    debug_print('RIGHT TREE')
+    debug_print('-' * 80)
+    right = c_parse_output(newfile)
+
+    c_compare_trees(left, right)
+
+
+#------------------------------------------------------------------------------
+
+
+# Register backends for specific file extensions.
+BACKENDS = \
+        {'.c': {'parse_output': c_parse_output, 'compare': c_compare},
+         '.h': {'parse_output': c_parse_output, 'compare': c_compare}}
+
+
+def get_backend(filename):
+    ''' Get an appropriate backend for FILENAME.
+    '''
+    name, ext = os.path.splitext(filename)
+
+    if not ext in BACKENDS.keys():
+        return None
+
+    return BACKENDS[ext]
+
+
+def exec_git_cmd(args):
+    ''' Execute a git command and return its result as a list of strings.
+    '''
+    args.insert(0, 'git')
+    debug_print(args)
+    proc = subprocess.Popen(args, stdout=subprocess.PIPE)
+
+    # Clean up the output by removing trailing spaces, newlines and dropping
+    # blank lines.
+    op = [decode(x[:-1]).strip() for x in proc.stdout]
+    op = [re.sub(r'[\s\f]+', ' ', x) for x in op]
+    op = [x for x in op if x]
+    return op
+
+
+def analyze_diff(oldfile, newfile, filename):
+    ''' Parse the output of the old and new files and print the difference.
+
+    For input files OLDFILE and NEWFILE with name FILENAME, generate reduced
+    trees for them and compare them.  We limit our comparison to only C source
+    files.
+    '''
+    backend = get_backend(filename)
+
+    if not backend:
+        return
+
+    backend['compare'](exec_git_cmd(['show', oldfile]),
+                       exec_git_cmd(['show', newfile]))
+
+
+IGNORE_LIST = [
+    'ChangeLog',
+    'sysdeps/x86_64/dl-trampoline.h'
+    ]
+
+
+def list_changes(commit):
+    ''' List changes in a single commit.
+
+    For the input commit id COMMIT, identify the files that have changed and the
+    nature of their changes.  Print commit information in the ChangeLog format,
+    calling into helper functions as necessary.
+    '''
+
+    op = exec_git_cmd(['show', '--date=short', '--raw', commit])
+    authors = []
+    date = ''
+    merge = False
+    copyright_exempt=''
+
+    for l in op:
+        if l.lower().find('copyright-paperwork-exempt:') == 0 \
+                and 'yes' in l.lower():
+            copyright_exempt=' (tiny change)'
+        elif l.lower().find('co-authored-by:') == 0 or \
+                l.find('Author:') == 0:
+            author = l.split(':')[1]
+            author = re.sub(r'(.*)\s*(<.*)', r'\1  \2', author)
+            authors.append(author)
+        elif l.find('Date:') == 0:
+            date=l[5:].strip()
+        elif l.find('Merge:') == 0:
+            merge = True
+
+    # Find raw commit information for all non-ChangeLog files.
+    op = [x[1:] for x in op if len(x) > 0 and re.match(r'^:[0-9]+', x)]
+
+    # Skip all ignored files.
+    for ign in IGNORE_LIST:
+        op = [x for x in op if ign not in x]
+
+    # It was only the ChangeLog, ignore.
+    if len(op) == 0:
+        return
+
+    print('%s  %s' % (date, authors[0]))
+
+    if (len(authors) > 1):
+        authors = authors[1:]
+        for author in authors:
+            print('            %s' % author)
+
+    print()
+
+    if merge:
+       print('\t MERGE COMMIT: %s\n' % commit)
+       return
+
+    print('\tCOMMIT%s: %s\n' % (copyright_exempt, commit))
+
+    # Each of these lines has a space separated format like so:
+    # :<OLD MODE> <NEW MODE> <OLD REF> <NEW REF> <OPERATION> <FILE1> <FILE2>
+    #
+    # where OPERATION can be one of the following:
+    # A: File added
+    # D: File removed
+    # M[0-9]{3}: File modified
+    # R[0-9]{3}: File renamed, with the 3 digit number following it indicating
+    # what percentage of the file is intact.
+    # C[0-9]{3}: File copied.  Same semantics as R.
+    # T: The permission bits of the file changed
+    # U: Unmerged.  We should not encounter this, so we ignore it/
+    # X, or anything else: Most likely a bug.  Report it.
+    #
+    # FILE2 is set only when OPERATION is R or C, to indicate the new file name.
+    #
+    # Also note that merge commits have a different format here, with three
+    # entries each for the modes and refs, but we don't bother with it for now.
+    #
+    # For more details: https://git-scm.com/docs/diff-format
+    for f in op:
+        data = f.split()
+        if data[4] == 'A':
+            print('\t* %s: New file.' % data[5])
+        elif data[4] == 'D':
+            print('\t* %s: Delete file.' % data[5])
+        elif data[4] == 'T':
+            print('\t* %s: Changed file permission bits from %s to %s' % \
+                    (data[5], data[0], data[1]))
+        elif data[4][0] == 'M':
+            print('\t* %s: Modified.' % data[5])
+            analyze_diff(data[2], data[3], data[5])
+        elif data[4][0] == 'R' or data[4][0] == 'C':
+            change = int(data[4][1:])
+            print('\t* %s: Move to...' % data[5])
+            print('\t* %s: ... here.' % data[6])
+            if change < 100:
+                analyze_diff(data[2], data[3], data[6])
+        # We should never encounter this, so ignore for now.
+        elif data[4] == 'U':
+            pass
+        else:
+            eprint('%s: Unknown line format %s' % (commit, data[4]))
+            sys.exit(42)
+
+    print('')
+
+
+def list_commits(revs):
+    ''' List commit IDs between the two revs in the REVS list.
+    '''
+    ref = revs[0] + '..' + revs[1]
+    return exec_git_cmd(['log', '--pretty=%H', ref])
+
+
+def main(revs):
+    ''' ChangeLog Generator Entry Point.
+    '''
+    commits = list_commits(revs)
+    for commit in commits:
+        list_changes(commit)
+
+
+def backend_file_test(f):
+    ''' Parser debugger Entry Point.
+    '''
+    backend = get_backend(f)
+
+    if not backend:
+        debug_print('%s: No backend for this file type, cannot debug' % f)
+        return
+
+    with open(f) as srcfile:
+        op = srcfile.readlines()
+        # Form feeds and line feeds removed.
+        op = [re.sub(r'\f+', r'', x[:-1]) for x in op]
+        tree = backend['parse_output'](op)
+
+
+# Program Entry point.  If -d is specified, the second argument is assumed to be
+# a file and only the backend for the file is run in verbose mode.
+if __name__ == '__main__':
+    if len(sys.argv) != 3:
+        usage(sys.argv[0])
+
+    if sys.argv[1] == '-d':
+        debug = True
+        backend_file_test(sys.argv[2])
+    else:
+        main(sys.argv[1:])