parsagon 0.17.0__tar.gz → 0.17.2__tar.gz

{parsagon-0.17.0 → parsagon-0.17.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: parsagon
-Version: 0.17.0
+Version: 0.17.2
 Summary: Allows you to create browser automations with natural language
 Author-email: Sandy Suh <sandy@parsagon.io>
 Project-URL: Homepage, https://parsagon.io
@@ -20,7 +20,9 @@ Requires-Dist: undetected-chromedriver==3.5.5
 Requires-Dist: webdriver-manager==4.0.2
 Requires-Dist: jsonpath-ng==1.5.3
 Requires-Dist: simplejson==3.19.1
-Requires-Dist: pypdf==4.2.0
+Requires-Dist: PyMuPDF==1.26.4
+Requires-Dist: pymupdf4llm==0.0.27
+Requires-Dist: pypdf==5.9.0
 Requires-Dist: python-dateutil==2.9.0.post0
 Requires-Dist: html2text==2024.2.26
 Requires-Dist: feedparser==6.0.11

{parsagon-0.17.0 → parsagon-0.17.2}/pyproject.toml

@@ -16,7 +16,7 @@ line-length = 120
 
 [project]
 name = "parsagon"
-version = "0.17.0"
+version = "0.17.2"
 description = "Allows you to create browser automations with natural language"
 readme = "README.md"
 requires-python = ">=3.8"
@@ -43,7 +43,9 @@ dependencies = [
     'webdriver-manager==4.0.2',
     'jsonpath-ng==1.5.3',
     'simplejson==3.19.1',
-    'pypdf==4.2.0',
+    'PyMuPDF==1.26.4',
+    'pymupdf4llm==0.0.27',
+    'pypdf==5.9.0',
     'python-dateutil==2.9.0.post0',
     'html2text==2024.2.26',
     'feedparser==6.0.11',

{parsagon-0.17.0 → parsagon-0.17.2}/src/parsagon/executor.py

@@ -16,6 +16,7 @@ import httpx
 from lxml import etree
 import lxml.html
 from lxml.html.clean import Cleaner
+import pymupdf4llm
 from pypdf import PdfReader
 from pyvirtualdisplay import Display
 import undetected_chromedriver as uc
@@ -706,6 +707,15 @@ class Executor:
         os.remove(most_recent_file)
         return text
 
+    def get_pdf_md(self, url):
+        window_id = self.goto(url)
+        self.close_window(window_id)
+        files = glob.glob("*")
+        most_recent_file = max(files, key=os.path.getmtime)
+        md_text = pymupdf4llm.to_markdown(most_recent_file)
+        os.remove(most_recent_file)
+        return md_text
+
     def str_to_iso8601(self, s):
         while s:
             try:

{parsagon-0.17.0 → parsagon-0.17.2}/src/parsagon.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: parsagon
-Version: 0.17.0
+Version: 0.17.2
 Summary: Allows you to create browser automations with natural language
 Author-email: Sandy Suh <sandy@parsagon.io>
 Project-URL: Homepage, https://parsagon.io
@@ -20,7 +20,9 @@ Requires-Dist: undetected-chromedriver==3.5.5
 Requires-Dist: webdriver-manager==4.0.2
 Requires-Dist: jsonpath-ng==1.5.3
 Requires-Dist: simplejson==3.19.1
-Requires-Dist: pypdf==4.2.0
+Requires-Dist: PyMuPDF==1.26.4
+Requires-Dist: pymupdf4llm==0.0.27
+Requires-Dist: pypdf==5.9.0
 Requires-Dist: python-dateutil==2.9.0.post0
 Requires-Dist: html2text==2024.2.26
 Requires-Dist: feedparser==6.0.11

{parsagon-0.17.0 → parsagon-0.17.2}/src/parsagon.egg-info/requires.txt

@@ -10,7 +10,9 @@ undetected-chromedriver==3.5.5
 webdriver-manager==4.0.2
 jsonpath-ng==1.5.3
 simplejson==3.19.1
-pypdf==4.2.0
+PyMuPDF==1.26.4
+pymupdf4llm==0.0.27
+pypdf==5.9.0
 python-dateutil==2.9.0.post0
 html2text==2024.2.26
 feedparser==6.0.11