Issue 7230 - python-abp line2dict vectorization

abp/filters/rpy.py

Index: abp/filters/rpy.py
===================================================================
--- a/abp/filters/rpy.py
+++ b/abp/filters/rpy.py
@@ -93,3 +93,32 @@
 
     """
     return strings2utf8(tuple2dict(parse_line(text, mode)))
+
+
+def lines2dict(string_list, mode='start'):
+    """Convert a list of filterlist strings to a dictionary.
+
+    All strings in the output dictionary will be UTF8 byte strings. This is
+    necessary to prevent unicode encoding errors in rPython conversion layer.
+
+    Parameters
+    ----------
+    string_list: iterable of str
+        Each string in the list can be a header, metadata, empty line,
+        include instruction, or filter.
+    mode: str
+        Parsing mode. Note: The default is 'start' in case any of the filters

[Vasily Kuznetsov, 2019/01/24 13:27:59]

I think the default should be 'body'. The data team are doing their own parsing
of header and metadata, so in their use case this function will likely only see
filter lines. Also this would be consistent with the behavior of `line2dict`,
which we want to "vectorize" here.

[rhowell, 2019/01/25 00:30:07]

Done.

+        are a header or metadata. For more information, see
+        `abp.filters.parser.parse_line`
+
+    Returns
+    -------
+    dict

[Vasily Kuznetsov, 2019/01/24 13:27:59]

If I understand it right, the output should be a list of dicts, not dict of
dicts. The idea of this function was to move the loop (line 122) from R to
Python to avoid going through the bridge at each iteration. The final product
that Stephan needs is a list again.

[rhowell, 2019/01/25 00:30:07]

Done.

+        With the parsing results and all strings converted to utf8 byte
+        strings.
+
+    """
+    result = {}
+    for string in string_list:
+        result[string] = line2dict(string, mode)
+    return result