diff --git a/pdfsearch.py b/pdfsearch.py
new file mode 100644
index 0000000..627a728
--- /dev/null
+++ b/pdfsearch.py
@@ -0,0 +1,55 @@
+import sys
+from pdfminer.pdfparser import PDFParser, PDFDocument
+from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter
+from pdfminer.converter import PDFPageAggregator
+from pdfminer.layout import LTTextBoxHorizontal, LAParams
+from pdfminer.pdfinterp import PDFTextExtractionNotAllowed
+import os
+
+
+def parse(path, keyword):
+    with open(path, 'rb') as fp:
+        parser = PDFParser(fp)
+        doc = PDFDocument()
+        parser.set_document(doc)
+        doc.set_parser(parser)
+        doc.initialize()
+
+    if not doc.is_extractable:
+        raise PDFTextExtractionNotAllowed
+    else:
+        res_man = PDFResourceManager()
+        lp = LAParams()
+        device = PDFPageAggregator(res_man, laparams=lp)
+        interpreter = PDFPageInterpreter(res_man, device)
+
+        index = 0
+        result = []
+        for page in doc.get_pages():
+            interpreter.process_page(page)
+            layout = device.get_result()
+            index += 1
+            for x in layout:
+                if isinstance(x, LTTextBoxHorizontal):
+                    text = x.get_text()
+                    if text.__contains__(keyword):
+                        result.append(index)
+    result = list(set(result))
+    result = sorted(result)
+    return result
+
+
+def file_name(file_dir):
+    files = []
+    for file in os.walk(file_dir):
+        files.append(file)   # os.walk()所在目录的所有非目录文件名
+    return files
+
+
+if __name__ == '__main__':
+    with open('./search.log', 'w') as f:
+        for i in file_name('./556')[0][2]:
+            log = parse('./556/'+i, 'KEAP1')
+            if log:
+                for j in log:
+                    f.write(i+' '+str(j)+'\n')