From b695d962d82c208396bcfc6e14d5f1ed048dae03 Mon Sep 17 00:00:00 2001
From: Stephen Mackenzie <stephen.mackenzie@methodstudios.com>
Date: Thu, 15 Oct 2015 16:06:12 -0400
Subject: [PATCH 1/2] Replaced newline-dependent CMS parser with more robust
 regex parser

---
 cdl_convert/parse.py | 20 +++++++++-----------
 1 file changed, 9 insertions(+), 11 deletions(-)

diff --git a/cdl_convert/parse.py b/cdl_convert/parse.py
index 1c59e86..7070ea8 100644
--- a/cdl_convert/parse.py
+++ b/cdl_convert/parse.py
@@ -469,10 +469,9 @@ def parse_cmx(input_file):  # pylint: disable=R0912,R0914
     cdls = []
 
     with open(input_file, 'rb') as edl:
-        lines = edl.readlines()
+        lines = '\n'.join(edl.readlines())
 
     filename = os.path.basename(input_file).split('.')[0]
-
     def parse_cmx_clip(cmx_tuple):
         """Parses a three line cmx clip tuple."""
         if len(cmx_tuple) != 3:
@@ -500,18 +499,17 @@ def parse_cmx_clip(cmx_tuple):
 
         return cc
 
-    for i, line in enumerate(lines):
-        if line != '\r\n':
-            # We only care about newlines when reading CMX, because
-            # we use those to kick off parsing the next take.
-            continue
-        if i + 3 <= len(lines):
-            cc = parse_cmx_clip(lines[i + 1:i + 4])
+    #This regex will avoid caring about extra stuff between the important lines we care about as long as the
+    #important lines we care about are in the right order
+    ccMatcher = re.compile(r'(\d*.+)(\n*.*)(\*FROM.+)(\n*.*)(\*ASC_SOP.+)(\n*.*)(\*ASC_SAT.+)(\n*.*)(\*SOURCE.+)')
+    clipEntries = ccMatcher.findall(lines)
+    for entry in clipEntries:
+        if entry[2] is not '' and entry[4] is not None and entry[6] is not None:
+            cc = parse_cmx_clip((entry[2],entry[4],entry[6]))
+            cdls.append(cc)
         else:
             continue
 
-        cdls.append(cc)
-
     ccc = collection.ColorCollection()
     ccc.file_in = input_file
     ccc.append_children(cdls)

From 6e9bc1f29499ca1d48e7004c2a90c2547f86d5c6 Mon Sep 17 00:00:00 2001
From: Stephen Mackenzie <stephen.mackenzie@methodstudios.com>
Date: Thu, 15 Oct 2015 17:45:14 -0400
Subject: [PATCH 2/2] adjustments to regex parsing in response to addtl edl
 examples

---
 cdl_convert/parse.py | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

diff --git a/cdl_convert/parse.py b/cdl_convert/parse.py
index 7070ea8..93ded1a 100644
--- a/cdl_convert/parse.py
+++ b/cdl_convert/parse.py
@@ -477,7 +477,7 @@ def parse_cmx_clip(cmx_tuple):
         if len(cmx_tuple) != 3:
             print(cmx_tuple)
             return
-        title = cmx_tuple[0].split()[1]
+        title = cmx_tuple[0].split(': ')[1]
 
         sop = re.match(
             r'^\*ASC_SOP \(([\d\. -]+)\)\(([\d\. -]+)\)\(([\d\. -]+)\)',
@@ -501,15 +501,26 @@ def parse_cmx_clip(cmx_tuple):
 
     #This regex will avoid caring about extra stuff between the important lines we care about as long as the
     #important lines we care about are in the right order
-    ccMatcher = re.compile(r'(\d*.+)(\n*.*)(\*FROM.+)(\n*.*)(\*ASC_SOP.+)(\n*.*)(\*ASC_SAT.+)(\n*.*)(\*SOURCE.+)')
+    ccMatcher = re.compile(r'(\d*.+)(\n*.*)(\*\ *FROM.+)(\n*.*)(\*\ *ASC_(SOP|SAT).+)(\n*.*)(\*\ *ASC_(SOP|SAT).+)')
     clipEntries = ccMatcher.findall(lines)
     for entry in clipEntries:
-        if entry[2] is not '' and entry[4] is not None and entry[6] is not None:
-            cc = parse_cmx_clip((entry[2],entry[4],entry[6]))
+        clip = entry[2]
+        sop = None
+        sat = None
+        i=0
+        for group in entry:
+            if group == 'SOP':
+                sop = entry[i-1]
+            if group == 'SAT':
+                sat = entry[i-1]
+            i += 1
+        if clip is not None and sop is not None and sat is not None:
+            cc = parse_cmx_clip((clip, sop, sat))
             cdls.append(cc)
         else:
             continue
 
+
     ccc = collection.ColorCollection()
     ccc.file_in = input_file
     ccc.append_children(cdls)