Fix #96 - properly handle @co tags in SAM files

CHANGES.md

@@ -1,6 +1,17 @@
 # Changes
 
-## 2.0.0-alpha.3 (dev)
+## 2.0.0-alpha.5 (2020.05.04)
+
+* Fix #97 (port from 1.1 branch)
+* Fix #96 - properly handle @CO tags in SAM files
+
+## 2.0.0-alpha.4 (2020.01.10)
+
+* bugfixes
+
+## 2.0.0-alpha.3 (2020.01.10)
+
+* bugfixes
 
 ## 2.0.0-alpha.2 (2020.01.02)
 
@@ -37,6 +48,14 @@
 * Enabled output to stdout by default with single-end and interleaved reads.
 * Migrated to setuptools_scm for version management.
 
+## v1.1.26 (2020.05.04)
+
+Fix #97 - handle single-end data in QC report
+
+## v1.1.25 (2020.01.10)
+
+Fix #71 - handle corrupted adapter cache file
+
 ## v1.1.24 (2019.11.17)
 
 * Fix #87 - Python 3.8 incompatibility - change time.clock() to time.process_time()

atropos/io/formatters.py

@@ -1,6 +1,6 @@
 from abc import ABCMeta, abstractmethod
 from pathlib import Path
-from typing import List, Optional, Tuple, Union
+from typing import List, Optional, Tuple, Union, cast
 
 from atropos.errors import UnknownFileTypeError
 from atropos.io import SequenceFileType
@@ -256,17 +256,20 @@ def _get_header(self) -> Optional[str]:
         if not self._header:
             return DEFAULT_SAM_HEADER
 
-        def create_row(_tags: dict):
-            tags_str = "\t".join(f"{key}:{val}" for key, val in _tags.items())
+        def create_row(_tags: Union[str, dict]):
+            if isinstance(_tags, dict):
+                tags_str = "\t".join(f"{key}:{val}" for key, val in _tags.items())
+            else:
+                tags_str = cast(str, _tags)
             return f"@{header_type}\t{tags_str}"
 
         rows = []
 
-        for header_type, tags_list in self._header.items():
-            if isinstance(tags_list, dict):
-                rows.append(create_row(tags_list))
+        for header_type, header_value in self._header.items():
+            if isinstance(header_value, dict):
+                rows.append(create_row(header_value))
             else:
-                rows.extend(create_row(tags) for tags in tags_list)
+                rows.extend(create_row(tags) for tags in header_value)
 
         return "\n".join(rows) + "\n"
 

atropos/io/readers/_impl.py

@@ -6,7 +6,7 @@
 from pathlib import Path
 import sys
 from typing import (
-    Callable, Iterator, IO, Optional, Sequence as SequenceType, Tuple, Type,
+    Callable, Dict, Iterator, IO, Optional, Sequence as SequenceType, Tuple, Type,
     Union, cast
 )
 
@@ -634,23 +634,30 @@ def __init__(self, sam_file: IO):
         self._header = {}
         self._header_size = 0
 
-        def add_header(fields: SequenceType[str]):
-            header_type = fields[0][1:]
-            tags = dict(
+        def fields_to_tags(fields: SequenceType[str]) -> Dict[str, str]:
+            return dict(
                 (pair[0], pair[1])
-                for pair in (kv.split(":") for kv in fields[1:])
+                for pair in (kv.split(":") for kv in fields)
             )
 
+        def add_header(fields: SequenceType[str]):
+            header_type = fields[0][1:]
             if header_type == "HD":
-                self._header["HD"] = tags
+                self._header["HD"] = fields_to_tags(fields[1:])
             else:
-                if header_type == "SQ" and "LN" in tags:
-                    tags["LN"] = int(tags["LN"])
-
                 if header_type not in self._header:
                     self._header[header_type] = []
 
-                self._header[header_type].append(tags)
+                if header_type == "CO":
+                    # the text after the first tab is free-form - even tab is allowed
+                    self._header[header_type].append("\t".join(fields[1:]))
+                else:
+                    tags = fields_to_tags(fields)
+
+                    if header_type == "SQ" and "LN" in tags:
+                        tags["LN"] = int(tags["LN"])
+
+                    self._header[header_type].append(tags)
 
         for line in self._reader:
             if line[0].startswith("@"):