Script to generate ChangeLog-like output from git log

Co-authored-by: Gabriel F. T. Gomes <gabriel@inconstante.net.br>
Reviewed-by: Gabriel F. T. Gomes <gabriel@inconstante.net.br>
Reviewed-by: Joseph Myers <joseph@codesourcery.com>

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.

Design
------

At a high level, the script analyses the raw output of a VCS, parses
the source files that have changed and attempts to determine what
changed.  The script driver needs three distinct components to be
fully functional for a repository:

- A vcstocl_quirks.py file that helps it parse weird patterns in
  sources that may result from preprocessor defines.
- A VCS plugin backend; the git backend is implemented for glibc
- A programming language parser plugin.  C is currently implemented.

Additional programming language parsers can be added to give more
detailed output for changes in those types of files.

For input in languages other than those that have a parser, the script
only identifies if a file has been added, removed, modified,
permissions changed, etc. but cannot understand the change in content.

The C Parser
------------

The C parser is capable of parsing C programs with preprocessor macros
in place, as if they were part of the language.  This presents some
challenges with parsing code that expands macros on the fly and to
help work around that, a vcstocl_quirks.py file has transformations to
ease things.

The C parser currently can identify macro definitions and scopes and
all global and static declarations and definitions.  It cannot parse
(and compare) changes inside functions yet, it could be a future
enhancement if the need for it arises.

Testing
-------

The script has been tested with the glibc repository up to glibc-2.29
and also in the past with emacs.  While it would be ideal to have
something like this in a repository like gnulib, that should not be a
bottleneck for glibc to start using this, so this patch proposes to
add these scripts into glibc.

And here is (hopefully!) one of the last ChangeLog entries we'd have
to write for glibc:

	* scripts/gitlog_to_changelog.py: New script to auto-generate
	ChangeLog.
	* scripts/vcs_to_changelog/frontend_c.py: New file.
	* scripts/vcs_to_changelog/misc_util.py: New file.
	* scripts/vcs_to_changelog/vcs_git.py: New file.
	* scripts/vcs_to_changelog/vcstocl_quirks.py: Likewise.

5 files changed, 1244 insertions, 0 deletions

diff --git a/scripts/gitlog_to_changelog.py b/scripts/gitlog_to_changelog.py
new file mode 100755
index 0000000000..41c47bdf74
--- /dev/null
+++ b/scripts/gitlog_to_changelog.py
@@ -0,0 +1,138 @@
+#!/usr/bin/python3
+# Main VCSToChangeLog script.
+# Copyright (C) 2019 Free Software Foundation, Inc.
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program 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 General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+''' Generate a ChangeLog style output based on a VCS log.
+
+This script takes two revisions as input and generates a ChangeLog style output
+for all revisions between the two revisions.
+
+This script is intended to be executed from the project parent directory.
+
+The vcs_to_changelog directory has a file vcstocl_quirks.py that defines a
+function called get_project_quirks that returns a object of class type
+ProjectQuirks or a subclass of the same.  The definition of the ProjectQuirks
+class is below and it specifies the properties that the project must set to
+ensure correct parsing of its contents.
+
+Among other things, ProjectQurks specifies the VCS to read from; the default is
+assumed to be git.  The script then studies the VCS log and for each change,
+list out the nature of changes in the constituent files.
+
+Each file type may have parser frontends that can read files and construct
+objects that may be compared to determine the minimal changes that occured in
+each revision.  For files that do not have parsers, we may only know the nature
+of changes at the top level depending on the information that the VCS stores.
+
+The parser frontend must have a compare() method that takes the old and new
+files as arrays of strings and prints the output in ChangeLog format.
+
+Currently implemented VCS:
+
+    git
+
+Currently implemented frontends:
+
+    C
+'''
+import sys
+import os
+import re
+import argparse
+from vcs_to_changelog.misc_util import *
+from vcs_to_changelog import frontend_c
+from vcs_to_changelog.vcs_git import *
+
+debug = DebugUtil(False)
+
+class ProjectQuirks:
+    # This is a list of regex substitutions for C/C++ macros that are known to
+    # break parsing of the C programs.  Each member of this list is a dict with
+    # the key 'orig' having the regex and 'sub' having the substitution of the
+    # regex.
+    MACRO_QUIRKS = []
+
+    # This is a list of macro definitions that are extensively used and are
+    # known to break parsing due to some characteristic, mainly the lack of a
+    # semicolon at the end.
+    C_MACROS = []
+
+    # The repo type, defaults to git.
+    repo = 'git'
+
+    # List of files to ignore either because they are not needed (such as the
+    # ChangeLog) or because they are non-parseable.  For example, glibc has a
+    # header file that is only assembly code, which breaks the C parser.
+    IGNORE_LIST = ['ChangeLog']
+
+
+# Load quirks file.  We assume that the script is run from the top level source
+# directory.
+sys.path.append('/'.join([os.getcwd(), 'scripts', 'vcs_to_changelog']))
+try:
+    from vcstocl_quirks import *
+    project_quirks = get_project_quirks(debug)
+except:
+    project_quirks = ProjectQuirks()
+
+def analyze_diff(filename, oldfile, newfile, frontends):
+    ''' 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.
+    '''
+    name, ext = os.path.splitext(filename)
+
+    if not ext in frontends.keys():
+        return None
+    else:
+        frontend = frontends[ext]
+        frontend.compare(oldfile, newfile)
+
+
+def main(repo, frontends, refs):
+    ''' ChangeLog Generator Entry Point.
+    '''
+    commits = repo.list_commits(args.refs)
+    for commit in commits:
+        repo.list_changes(commit, frontends)
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument('refs', metavar='ref', type=str, nargs=2,
+                     help='Refs to print ChangeLog entries between')
+
+    parser.add_argument('-d', '--debug', required=False, action='store_true',
+                     help='Run the file parser debugger.')
+
+    args = parser.parse_args()
+
+    debug.debug = args.debug
+
+    if len(args.refs) < 2:
+        debug.eprint('Two refs needed to get a ChangeLog.')
+        sys.exit(os.EX_USAGE)
+
+    REPO = {'git': GitRepo(project_quirks.IGNORE_LIST, debug)}
+
+    fe_c = frontend_c.Frontend(project_quirks, debug)
+    FRONTENDS = {'.c': fe_c,
+                 '.h': fe_c}
+
+    main(REPO[project_quirks.repo], FRONTENDS, args.refs)
diff --git a/scripts/vcs_to_changelog/frontend_c.py b/scripts/vcs_to_changelog/frontend_c.py
new file mode 100644
index 0000000000..4191e86456
--- /dev/null
+++ b/scripts/vcs_to_changelog/frontend_c.py
@@ -0,0 +1,827 @@
+#!/usr/bin/python3
+# The C Parser.
+# Copyright (C) 2019 Free Software Foundation, Inc.
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program 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 General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+from enum import Enum
+import re
+from vcs_to_changelog.misc_util import *
+
+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'}}
+
+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
+    block['actions'] = actions[type]
+
+    return block
+
+
+class ExprParser:
+    ''' Parent class of all of the C expression parsers.
+
+    It is necessary that the children override the parse_line() method.
+    '''
+    ATTRIBUTE = r'(((__attribute__\s*\(\([^;]+\)\))|(asm\s*\([?)]+\)))\s*)*'
+
+    def __init__(self, project_quirks, debug):
+        self.project_quirks = project_quirks
+        self.debug = debug
+
+    def fast_forward_scope(self, 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_line(self, cur, op, loc, code, macros):
+        ''' The parse method should always be overridden by the child.
+        '''
+        raise
+
+
+class FuncParser(ExprParser):
+    REGEX = re.compile(ExprParser.ATTRIBUTE + r'\s*(\w+)\s*\([^(][^{]+\)\s*{')
+
+    def parse_line(self, cur, op, loc, code, macros):
+        ''' Parse a function.
+
+        Match a function definition.
+
+        - 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
+
+        - Returns: The next location to be read in the array.
+        '''
+        found = re.search(self.REGEX, cur)
+        if not found:
+            return cur, loc
+
+        name = found.group(5)
+        self.debug.print('FOUND FUNC: %s' % name)
+
+        # Consume everything up to the ending brace of the function.
+        (cur, loc) = self.fast_forward_scope(cur, op, loc)
+
+        new_block(name, block_type.func, [cur], code)
+
+        return '', loc
+
+
+class CompositeParser(ExprParser):
+    # Composite types such as structs and unions.
+    REGEX = re.compile(r'(struct|union|enum)\s*(\w*)\s*{')
+
+    def parse_line(self, cur, op, loc, code, macros):
+        ''' Parse a composite type.
+
+        Match declaration of a composite type such as a sruct or a union..
+
+        - 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
+
+        - Returns: The next location to be read in the array.
+        '''
+        found = re.search(self.REGEX, cur)
+        if not found:
+            return cur, loc
+
+        # Lap up all of the struct definition.
+        (cur, loc) = self.fast_forward_scope(cur, op, loc)
+
+        name = found.group(2)
+
+        if not name:
+            if 'typedef' in cur:
+                name = re.sub(r'.*}\s*(\w+);$', r'\1', cur)
+            else:
+                name= '<anoymous>'
+
+        ctype = found.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, block_type.composite, [cur], code)
+
+        return '', loc
+
+
+class AssignParser(ExprParser):
+    # Static assignments.
+    REGEX = re.compile(r'(\w+)\s*(\[[^\]]*\])*\s*([^\s]*attribute[\s\w()]+)?\s*=')
+
+    def parse_line(self, cur, op, loc, code, macros):
+        ''' Parse an assignment statement.
+
+        This includes array assignments.
+
+        - 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
+
+        - Returns: The next location to be read in the array.
+        '''
+        found = re.search(self.REGEX, cur)
+        if not found:
+            return cur, loc
+
+        name = found.group(1)
+        self.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, block_type.assign, [cur], code)
+
+        return '', loc
+
+
+class DeclParser(ExprParser):
+    # Function pointer typedefs.
+    TYPEDEF_FN_RE = re.compile(r'\(\*(\w+)\)\s*\([^)]+\);')
+
+    # Simple decls.
+    DECL_RE = re.compile(r'(\w+)(\[\w*\])*\s*' + ExprParser.ATTRIBUTE + ';')
+
+    # __typeof decls.
+    TYPEOF_RE = re.compile(r'__typeof\s*\([\w\s]+\)\s*(\w+)\s*' + \
+                           ExprParser.ATTRIBUTE + ';')
+
+    # Function Declarations.
+    FNDECL_RE = re.compile(r'\s*(\w+)\s*\([^\(][^;]*\)\s*' +
+                           ExprParser.ATTRIBUTE + ';')
+
+    def __init__(self, regex, blocktype, project_quirks, debug):
+        # The regex for the current instance.
+        self.REGEX = regex
+        self.blocktype = blocktype
+        super().__init__(project_quirks, debug)
+
+    def parse_line(self, cur, op, loc, code, macros):
+        ''' Parse a top level declaration.
+
+        All types of declarations except function declarations.
+
+        - 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
+
+        - Returns: The next location to be read in the array.
+        '''
+        found = re.search(self.REGEX, cur)
+        if not found:
+            return cur, loc
+
+        # The name is the first group for all of the above regexes.  This is a
+        # coincidence, so care must be taken if regexes are added or changed to
+        # ensure that this is true.
+        name = found.group(1)
+
+        self.debug.print('FOUND DECL: %s' % name)
+        new_block(name, self.blocktype, [cur], code)
+
+        return '', loc
+
+
+class MacroParser(ExprParser):
+    # 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*(\(.*\))*$')
+
+    def parse_line(self, cur, op, loc, code, macros):
+        ''' 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
+        - MACROS 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(self.MACROCALL_RE, cur)
+        if found:
+            sym = found.group(1)
+            name = found.group(2)
+            if sym in macros or self.project_quirks and \
+                    sym in self.project_quirks.C_MACROS:
+                self.debug.print('FOUND MACROCALL: %s (%s)' % (sym, name))
+                new_block(sym, block_type.macrocall, [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():
+            self.debug.print('FOUND MACROCALL (without brackets): %s' % (cur))
+            new_block(cur, block_type.macrocall, [cur], code)
+            return '',loc
+
+        return cur, loc
+
+
+class Frontend:
+    ''' The C Frontend implementation.
+    '''
+    KNOWN_MACROS = []
+
+    def __init__(self, project_quirks, debug):
+        self.op = []
+        self.debug = debug
+        self.project_quirks = project_quirks
+
+        self.c_expr_parsers = [
+                CompositeParser(project_quirks, debug),
+                AssignParser(project_quirks, debug),
+                DeclParser(DeclParser.TYPEOF_RE, block_type.decl,
+                           project_quirks, debug),
+                DeclParser(DeclParser.TYPEDEF_FN_RE, block_type.decl,
+                           project_quirks, debug),
+                DeclParser(DeclParser.FNDECL_RE, block_type.fndecl,
+                           project_quirks, debug),
+                FuncParser(project_quirks, debug),
+                DeclParser(DeclParser.DECL_RE, block_type.decl, project_quirks,
+                           debug),
+                MacroParser(project_quirks, debug)]
+
+
+    def remove_extern_c(self):
+        ''' 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 self.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.
+        self.op = new_op
+        new_op = []
+        nesting = 0
+        extern_nesting = 0
+        in_cpp = False
+        for l in self.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)
+
+        self.op = new_op
+
+
+    def remove_comments(self, op):
+        ''' Remove comments.
+
+        Return OP by removing all comments from it.
+        '''
+        self.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)))
+
+