Skip to content
Snippets Groups Projects
Commit bea58828 authored by skurzinz's avatar skurzinz
Browse files

use real txt filename in the zipped file contents

parent c744aba2
No related branches found
No related tags found
1 merge request!1use real txt filename in the zipped file contents
Show whitespace changes
Inline Side-by-side
Showing
with 2 additions and 3 deletions
txtDownloader.ipynb
+ 2
− 3
View file @ bea58828
%% Cell type:code id: tags:
``` python
import urllib.request, json, shutil, os
from zipfile import ZipFile
from IPython.display import FileLink
```
%% Cell type:markdown id: tags:
e.g.:+Z196807705
%% Cell type:code id: tags:
``` python
barcode=input()
```
%% Output
+Z196807705
%% Cell type:code id: tags:
``` python
url="https://iiif.onb.ac.at/presentation/ABO/"+barcode+"/manifest/"
url
```
%% Output
'https://iiif.onb.ac.at/presentation/ABO/+Z196807705/manifest/'
%% Cell type:code id: tags:
``` python
response = urllib.request.urlopen(url)
data = json.loads(response.read())
i=0
zipObj = ZipFile(barcode + ".zip", "w")
all_txt = open("all.txt","wb")
for page in data["sequences"][0]["canvases"]:
txt_url = page["otherContent"][0]["resources"][0]["resource"]["@id"]
i=i+1
print("downloading " + txt_url)
output_file = str(i) + ".txt"
output_file = str(txt_url.split('/')[-1])
with urllib.request.urlopen(txt_url) as response, open(output_file, "wb") as out_file:
shutil.copyfileobj(response, out_file)
zipObj.write(os.path.abspath(output_file),output_file)
txt_page = open(output_file, "rb")
shutil.copyfileobj(txt_page, all_txt)
txt_page.close
os.remove(output_file)
zipObj.write("all.txt")
all_txt.close()
os.remove("all.txt")
zipObj.close()
FileLink(barcode + ".zip")
```
%% Output
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000001.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000002.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000003.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000004.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000005.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000006.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000007.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000008.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000009.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000010.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000011.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000012.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000013.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000014.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000015.txt
downloading https://iiif.onb.ac.at/presentation/ABO/Z196807705/resource/00000016.txt
/home/georg/Projekte/sacha/sacha-txt-downloader/+Z196807705.zip
%% Cell type:code id: tags:
``` python
```
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment