from __future__ import generators

import os
import re
import time

_date_pat = re.compile('^(\d\d\d\d)-(\d\d)-(\d\d)\s+(.*)\s*$')

def _parse_header(line):
    if line[0].isspace(): return None
    match = _date_pat.match(line)
    if match:
        when = (int(match.group(1)),
                int(match.group(2)),
                int(match.group(3)),
                0,0,0, 0,0,0)
        who = match.group(4)
        return (when, who)
    parts = line.split(None, 5)
    if len(parts) == 6:
        try:
            timeval = ' '.join(parts[:5])
            when = time.strptime(timeval, '%a %b %d %H:%M:%S %Y')
            who = parts[5]
            return (when, who)
        except ValueError:
            pass
    return None

def parse(fp):
    '''parse a changelog in the standard format

    Yields tuples of the form (when, who, changes).'''
    when = None
    who = None
    changes = []

    for line in fp:
        header = _parse_header(line)
        if header:
            if when:
                yield (when, who, changes)
            when, who = header
            changes = []
        else:
            changes.append(line)
    if when:
        yield (when, who, changes)

def _find_changelogs(checkoutdir, base):
    if checkoutdir[-1] != '/': checkoutdir = checkoutdir + '/'
    def walker(changelogs, dirname, fnames):
        if 'ChangeLog' in fnames:
            changelogs.append(os.path.join(dirname, 'ChangeLog'))
        elif 'ChangeLog,v' in fnames:
            changelogs.append(os.path.join(dirname, 'ChangeLog,v'))
    changelogs = []
    os.path.walk(os.path.join(checkoutdir, base), walker, changelogs)
    for i in range(len(changelogs)):
        changelogs[i] = changelogs[i][len(checkoutdir):]
    return changelogs

def aggregate(checkoutdir, base):
    filenames = _find_changelogs(checkoutdir, base)
    logs = []
    for i in range(len(filenames)):
        filename = os.path.join(checkoutdir, filenames[i])
        if filename.endswith(',v'):
            fp = os.popen('co -p %s 2>/dev/null' % filename, 'r')
        else:
            fp = file(filename, 'r')
        parser = parse(fp)
        try:
            item = parser.next()
            logs.append((filenames[i], parser, item))
        except StopIteration:
            pass

    def compare_tuples(t1, t2):
        '''compare two (filename, parser, item) tuples by date'''
        item1 = t1[2]
        item2 = t2[2]
        return cmp(item1[0], item2[0])

    while logs:
        logs.sort(compare_tuples)  # sort logs by time of last read entry
        (filename, parser, item) = logs[-1]
        del logs[-1]
        yield (filename,) + item
        # now read another entry from that log
        try:
            item = parser.next()
            logs.append((filename, parser, item))
        except StopIteration:
            pass

def _escape(s):
    s = s.replace('&', '&')
    s = s.replace('<', '&lt;')
    s = s.replace('>', '&gt;')
    return s

def _break_into_paragraphs(lines):
    paragraphs = [[]]
    for line in lines:
        line = _escape(line.strip())
        if not line:
            if paragraphs[-1]:
                paragraphs.append([])
            continue
        if line[0] == '*': # bullets start new paragraphs
            if paragraphs[-1]:
                paragraphs.append([])
        paragraphs[-1].append(line)
    if not paragraphs[-1]: del paragraphs[-1]
    return paragraphs

def _splitword(s):
    parts = s.split(None, 1)
    parts.append('')
    return parts[0], parts[1]

email_pat = re.compile(r'(&lt;(.*@[-a-zA-Z0-9.]+)&gt;)')
bug_pat = re.compile(r'((?:bug\s+#?|#)(\d\d\d\d*))', re.I)

def _linkify(s, bugsroot=None):
    s = email_pat.sub(r'<a href="mailto:\2">\1</a>', s)
    if bugsroot:
        s = bug_pat.sub(r'<a href="%s\2">\1</a>' % bugsroot, s)
    return s

function_pat = re.compile(r'^(\s*\()([^\)]+)(\):.*)$')

def changes_to_html(changes, webcvsroot=None, bugsroot=None):
    ret = [ '<ul>\n' ]
    paragraphs = _break_into_paragraphs(changes)

    for para in paragraphs:
        if para[0][0] == '*':
            ret.append('<li>\n')
            # handle first line ...
            line = para[0][1:]
            while line:
                word, rest = _splitword(line)
                if word[0] == '(':
                    break
                line = rest
                trailer = ''
                if word[-1] in ',:':
                    trailer = word[-1]
                    word = word[:-1]
                if webcvsroot:
                    ret.append('<a href="%s%s">%s</a>'
                               % (webcvsroot, word, word))
                else:
                    ret.append('<b>%s</b>' % word)
                ret.append(trailer)
                ret.append(' ')
                if trailer == ':': break
            line = function_pat.sub(r'\1<i>\2</i>\3', line)
            ret.append(_linkify(line, bugsroot) + '\n')

            # print rest of lines
            for line in para[1:]:
                if function_pat.match(line):
                    ret.append('<br />\n')
                    line = function_pat.sub(r'\1<i>\2</i>\3', line)
                ret.append(_linkify(line, bugsroot) + '\n')
            ret.append('</li>\n')
        else:
            ret.append('<p>\n')
            for line in para:
                ret.append(_linkify(line, bugsroot) + '\n')
            ret.append('</p>\n')
    ret.append('</ul>\n')
    return ''.join(ret)