added gtts and readability

3 files changed, 11 insertions, 3 deletions

diff --git a/main.py b/main.py
index e5794e1..c72b674 100755
--- a/main.py
+++ b/main.py
@@ -9,6 +9,9 @@ from contextlib import closing
 from requests import get
 from requests.exceptions import RequestException
 from re import findall
+from readability import Document
+from gtts import gTTS
+from datetime import datetime as time
 
 
 class Argparser(object):
@@ -85,8 +88,11 @@ def main():
             try:
                 a.download()
                 a.parse()
-                # print(a.html)
-                print(a.text)
+                doc = Document(a.html)
+                print(doc.summary())
+                if a.text != '':
+                    tts = gTTS(a.text)
+                    tts.save(time.today().strftime("%b-%d-%Y-%M-%S-%f")+".mp3")
             except Exception as e:
                 logging.exception(e)
 
diff --git a/pyproject.toml b/pyproject.toml
index f91a38f..8eaddf2 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -9,6 +9,8 @@ license = "GPL3.0"
 python = "^3.8"
 newspaper3k = "^0.2.8"
 beautifulsoup4 = "^4.9.3"
+readability-lxml = "^0.8.1"
+gtts = "^2.2.3"
 
 [tool.poetry.dev-dependencies]
 
diff --git a/run.sh b/run.sh
index 710ec7f..be3cd55 100755
--- a/run.sh
+++ b/run.sh
@@ -3,7 +3,7 @@ set -e
 set -x
 
 # sniff --src https://github.com/coinpride/CryptoList --url | ./main.py > out2.txt
-./main.py --source https://github.com/coinpride/CryptoList > out.txt
+./main.py --source https://github.com/coinpride/CryptoList > out.html
 # ./main.py --source https://techurls.com/ > out.txt
 # cat out2.txt| pico2wave --wave=out2.wav
 # ./main.py --source http://blog.terminaldweller.com > out.txt