aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMartin Vilcans <martin@librador.com>2012-01-14 01:45:19 +0100
committerMartin Vilcans <martin@librador.com>2012-01-14 01:47:41 +0100
commit6466690e597a27597fa6dcf93f7b9d7035479471 (patch)
tree290918a78a685a01e7d51ccebcd25ee7e13cf6a2
parent8485c49836ea746016101e6fc3f941564ced75c3 (diff)
downloadscreenplain-6466690e597a27597fa6dcf93f7b9d7035479471.tar.gz
Ignore trailing \r and \n properly.
This caused the "two spaces creates blank line inside dialog" trick to fail.
-rw-r--r--screenplain/export/pdf.py1
-rw-r--r--screenplain/parsers/spmd.py51
2 files changed, 34 insertions, 18 deletions
diff --git a/screenplain/export/pdf.py b/screenplain/export/pdf.py
index 3e9fa70..4e29f6d 100644
--- a/screenplain/export/pdf.py
+++ b/screenplain/export/pdf.py
@@ -2,7 +2,6 @@
# Licensed under the MIT license:
# http://www.opensource.org/licenses/mit-license.php
-#!/usr/bin/python
import fileinput
from reportlab.pdfgen import canvas
from reportlab.lib import pagesizes
diff --git a/screenplain/parsers/spmd.py b/screenplain/parsers/spmd.py
index 7bdb79b..afa2717 100644
--- a/screenplain/parsers/spmd.py
+++ b/screenplain/parsers/spmd.py
@@ -24,10 +24,7 @@ slug_prefixes = (
TWOSPACE = ' ' * 2
centered_re = re.compile(r'\s*>\s*(.*)\s*<\s*$')
-
-
-def is_blank(string):
- return string == '' or string.isspace() and string != TWOSPACE
+preprocess_re = re.compile(r'^(\s*)(.*?)([ \t]*)[\r\n]*$')
def is_slug(blanks_before, line_list):
@@ -94,29 +91,49 @@ def create_paragraph(blanks_before, line_list):
return Action(_to_rich(line_list))
-def clean_line(line):
- """Strips leading whitespace and trailing end of line characters
- in a string.
+def _preprocess_line(raw_line):
+ """Splits a line into leading spaces, text content, and trailing spaces.
+
+ >>> _preprocess_line(' foo ')
+ (' ', 'foo', ' ')
- Leading whitespace is insignificant in SPMD, and trailing EOL
- appear when reading from a file or HTML form.
+ For a blank line, the trailing spaces will be returned as trailing
+ whitespace:
+
+ >>> _preprocess_line(' ')
+ ('', '', ' ')
"""
- stripped = line.rstrip('\r\n')
- if line == ' ':
- return line
- return line.lstrip()
+ line = raw_line.expandtabs(4)
+ leading, text, trailing = preprocess_re.match(line).groups()
+ if not text:
+ trailing = leading
+ leading = ''
+ return leading, text, trailing
+
+
+def _is_blank(preprocessed_line):
+ leading, text, trailing = preprocessed_line
+ return not text and not trailing
def parse(source):
"""Reads raw text input and generates paragraph objects."""
blank_count = 0
- source = (clean_line(line) for line in source)
+ source = (_preprocess_line(line) for line in source)
paragraphs = []
- for blank, lines in itertools.groupby(source, is_blank):
+ for blank, preprocessed_lines in itertools.groupby(source, _is_blank):
if blank:
- blank_count = len(list(lines))
+ blank_count = sum(1 for line in preprocessed_lines)
else:
- paragraphs.append(create_paragraph(blank_count, list(lines)))
+ paragraphs.append(
+ create_paragraph(
+ blank_count,
+ [
+ text + trailing
+ for (leading, text, trailing) in preprocessed_lines
+ ]
+ )
+ )
# Make a second pass over the script and replace transitions not
# followed by sluglines with action.