A couple of fixes on write_xml in create_xml.py + Created form_csv.py script

2020-12-08 08:01:17 +01:00
parent c18aaff11f
commit 75b015dcda
3 changed files with 207 additions and 34 deletions
@@ -0,0 +1,68 @@
+import argparse
+import csv
+import os
+
+from lxml import etree
+
+
+def write_general_statistics(path, out_list):
+    if len(out_list) == 0:
+        return
+    with open(path, 'w') as csvfile:
+        writer = csv.writer(csvfile, delimiter='\t',
+                                quotechar='"')
+        writer.writerow(['semanticRole', 'valency_pattern_ratio', 'valency_sentence_ratio'])
+        for line in out_list:
+            writer.writerow(line)
+
+
+def main(args):
+    for file in sorted(os.listdir(args.input)):
+        path = os.path.join(args.input, file)
+        tree = etree.parse(path)
+        gf_output = []
+        ssj_output = []
+        head = next(tree.iter('head'))
+        headword = head.find('headword').find('lemma').text
+        #for div in root.iterfind('.//div'):
+        for elem in tree.iter('statisticsContainer'):
+        # for element in tree.iterfind('statisticsContainer'):
+        # for element in tree.find('statisticsContainer'):
+            semRole = elem.find('semanticRole').text
+            gf_pattern = None
+            gf_sentence = None
+            ssj_pattern = None
+            ssj_sentence = None
+            measure = elem.find('measureList')
+            for el in measure:
+                if el.attrib['type'] == 'valency_pattern_ratio' and el.attrib['source'] == 'Gigafida 2.0':
+                    gf_pattern = el.text
+                if el.attrib['type'] == 'valency_sentence_ratio' and el.attrib['source'] == 'Gigafida 2.0':
+                    gf_sentence = el.text
+                if el.attrib['type'] == 'valency_pattern_ratio' and el.attrib['source'] == 'ssj500k 2.2':
+                    ssj_pattern = el.text
+                if el.attrib['type'] == 'valency_sentence_ratio' and el.attrib['source'] == 'ssj500k 2.2':
+                    ssj_sentence = el.text
+            if gf_pattern is not None and gf_sentence is not None:
+                gf_output.append([semRole, gf_pattern, gf_sentence])
+            if ssj_pattern is not None and ssj_sentence is not None:
+                ssj_output.append([semRole, ssj_pattern, ssj_sentence])
+
+        for elem in tree.iter('valencyPattern'):
+            valency_pattern_id = elem.attrib['id']
+            measure = None
+            for measure_el in elem.find('measure'):
+                if measure_el.attrib['source'] == 'Gigafida 2.0':
+                    measure = measure_el.text
+
+        write_general_statistics(os.path.join(args.output, headword + '_gf_stats.tsv'), gf_output)
+        write_general_statistics(os.path.join(args.output, headword + '_ssj_stats.tsv'), ssj_output)
+
+if __name__ == '__main__':
+    arg_parser = argparse.ArgumentParser(description='Export and validate collocation data from DDD database.')
+    arg_parser.add_argument('--input', type=str, help='Input directory')
+    arg_parser.add_argument('--output', type=str, help='Output directory')
+
+    args = arg_parser.parse_args()
+
+    main(args)