Thank you so much for the help this works an absolute treat
Just out of curiosity, were you using my script, or ccstoneâs?
I think theyâre pretty similar except heâs doing everything in the shell and Iâm using as much Applescript as possible and just using the shell to call curl to do the actual download.
If you know some Applescript but not much shell, then mine might have some modification/maintainability advantage for you⊠of course, if you know shell better then Applescript, then the opposite would be true.
Hey guys,
I keep getting an error when using these URLs, I have a list of over 300 and I have changed the base URL.
http://jpl.a.bigcontent.io/v1/static/bl_006773_a
http://jpl.a.bigcontent.io/v1/static/bl_006806_a
http://jpl.a.bigcontent.io/v1/static/bl_006908_a
Not sure why? Can anyone help?
Hey Eastland,
Whose script are you using?
Whatâs the base-url in your script?
What exactly is the error?
-Chris
Hey ccstone,
Thanks for having a look, I am using t.spoons script
use AppleScript version "2.4" -- Yosemite (10.10) or later
use scripting additions
set baseURL to "http://jpl.a.bigcontent.io/v1/static/"
set theSpreadsheet to choose file with prompt "Please select your csv file with the image URL's" of type {"csv"} without multiple selections allowed
tell application "Finder"
set downloadsFolder to path to downloads folder
set saveFolder to ((downloadsFolder as text) & "Script Downloaded Images:")
if not (exists saveFolder) then make new folder at downloadsFolder with properties {name:"Script Downloaded Images"}
set saveFolder to POSIX path of saveFolder
end tell
set spreadsheetData to read theSpreadsheet
set dataLines to the paragraphs of spreadsheetData
set theURLs to {}
repeat with aLine in dataLines
set delimitHolder to AppleScript's text item delimiters
set AppleScript's text item delimiters to ","
set newURL to baseURL & text item 1 of aLine
copy newURL to the end of theURLs
set AppleScript's text item delimiters to delimitHolder
end repeat
set imageCount to the count of theURLs
set missingImages to {}
set downloadedImages to {}
repeat with imageURL in theURLs
set delimitHolder to AppleScript's text item delimiters
set AppleScript's text item delimiters to "/"
set imageName to the last text item of imageURL
set AppleScript's text item delimiters to delimitHolder
set downloadFilePath to saveFolder & imageName as text
set shellOutput to (do shell script "
image=\"" & (contents of imageURL as text) & "\"
response=$(curl --write-out %{http_code} --silent --output /dev/null $image)
if [ \"$response\" == \"200\" ]; then
curl $image -o \"" & downloadFilePath & "\"
echo \"Image downloaded\"
else
echo \"Image does not exist\"
fi
")
if shellOutput is "Image does not exist" then
copy imageURL to the end of missingImages
else
copy imageURL to the end of downloadedImages
end if
end repeat
set missingImageDialogText to ""
if (count of missingImages) > 0 then
set missingImageDialogText to "The following images failed to download:"
repeat with anImage in missingImages
set missingImageDialogText to missingImageDialogText & return & anImage
end repeat
end if
display dialog "The script completed." & return & (count of downloadedImages) & " out of " & imageCount & " images were successfully downloaded to: " & saveFolder & return & missingImageDialogText buttons {"Cancel", "OK"} default button "OK"
I am using the base URL http://jpl.a.bigcontent.io/v1/static/
and getting this error
The script completed.
0 out of 152 images were successfully downloaded to: /Users/eastland/Downloads/Script Downloaded Images/
The following images failed to download:
http://jpl.a.bigcontent.io/v1/static/ĂâÂȘĂžbl_006773_a
http://jpl.a.bigcontent.io/v1/static/bl_006806_a
http://jpl.a.bigcontent.io/v1/static/bl_006908_a
http://jpl.a.bigcontent.io/v1/static/bl_007373_a
http://jpl.a.bigcontent.io/v1/static/bl_012337_a
and so onâŠ
I have checked the URLs and some auto download the mage when entered into a browser not sure if this may cause issue.
Can you help?
I opened my script from your post, with the base URL modified as you did it, put the URLâs you posted (with the base path stripped back off) into my test .csv:
And my result was that 4 out of five downloaded successfully.
Only the first one failed, and it also fails if I put the URL you provided into a browser.
Those weird characters in the first link make me suspect a possible problem with character encoding. That the values in your CSV, unlike all but your first posted value, are getting characters changed between what youâre seeing the CSV and what terminal is actually trying to retrieve. For example, terminal doesnât use Unicode text.
But thatâs just a guess, and I never have to mess with text encoding problems (yet), so Iâm really not sure.
But for me, the script is working perfectly with the URLâs you posted, aside from the first that doesnât work in a browser either.
If you want me to look more, you could PM a link to download an actual sample of a CSV youâre using for these, and I can try to dig a little deeper.
Thanks t.spoon,
I think I maybe saving out from excel incorrectly I have PMâd you a link to the file, thanks for having a look.
Well, this just got hard for me to help you with. I opened the script from your last post, downloaded the .csv you PMâd me, ran the script, and got:
The script completed.
146 out of 146 images were successfully downloaded to: ~/Downloads/Script Downloaded Images/
So being unable to reproduce the error, itâs hard to troubleshoot.
Are you clearing out the folder between runs? Iâm not entirely sure what the shell script does in case of name conflicts with already downloaded files. If clearing out the folder helps, I can easily modify my script to be like ccstoneâs and create a new folder for each run.
But I sort of doubt thatâs the problem, and Iâm afraid Iâm short of ideas on what the problem could be. It really does get hard when the error isnât reproducible.
Do the files download on your computer when you manually put the URL into a browser?
- Tom.
Thanks for having a look into this for me, I cleared all the folders and re ran the script with no luck still getting 0 of 146 images were downloaded. I then copied individual URLs into my browser the first two auto downloaded the third showed the image in the browser.
I connect through a proxy do you think this could be an issue?
The odd thing is if I revert back to using the older base URL http://i1.adis.ws/i/jpl/ this works fine but using the new URL http://jpl.a.bigcontent.io/v1/static/ still not working
The reason for changing URL is the image file returned has less compression which is the one I need.
Yeah, I donât know. I double-checked, and it is working fine for me with the base URL set to
It does seem particularly odd that the script doesnât work for you with that base URL but does work for me with the same, but for other Base URLâs it works for both of us.
Iâm just scratching my head here.
- Tom.
Weâre both using similar âdo shell scriptâ lines with CURL for the actualy downloading, so the results will probably be the same if you use ccstoneâs script.
But since what weâre looking at seems to be a weird, inconsistent bug, Iâd give his script a shot and see if you get a different result.
- Tom.
Hi Tom,
Yeah iâm confused with this one, I tried CCstones script but again it returned errors, it only brought back the last cell and skipped the rest.
If the script was modified to just read the CSV file with full links in the doc do you think this may work? Basically ditch the base URL function from the script.
Or am I clutching at straws?
Probably clutching at straws, but who knows. Computers can be weird.
Here it is expecting the full URL in the file.
I tested it with a few files you provided using the âhttp://jpl.a.bigcontent.io/v1/static/â URL beginning and the full URL in the .csv and it worked fine.
- Tom.
use AppleScript version "2.4" -- Yosemite (10.10) or later
use scripting additions
set theSpreadsheet to choose file with prompt "Please select your csv file with the image URL's" of type {"csv"} without multiple selections allowed
tell application "Finder"
set downloadsFolder to path to downloads folder
set saveFolder to ((downloadsFolder as text) & "Script Downloaded Images:")
if not (exists saveFolder) then make new folder at downloadsFolder with properties {name:"Script Downloaded Images"}
set saveFolder to POSIX path of saveFolder
end tell
set spreadsheetData to read theSpreadsheet
set dataLines to the paragraphs of spreadsheetData
set theURLs to {}
set delimitHolder to AppleScript's text item delimiters
repeat with aLine in dataLines
set AppleScript's text item delimiters to ","
copy text item 1 of aLine to the end of theURLs
end repeat
set AppleScript's text item delimiters to delimitHolder
set imageCount to the count of theURLs
set missingImages to {}
set downloadedImages to {}
set delimitHolder to AppleScript's text item delimiters
repeat with imageURL in theURLs
set AppleScript's text item delimiters to "/"
set imageName to the last text item of imageURL
set downloadFilePath to saveFolder & imageName as text
set shellOutput to (do shell script "
image=\"" & (contents of imageURL as text) & "\"
response=$(curl --write-out %{http_code} --silent --output /dev/null $image)
if [ \"$response\" == \"200\" ]; then
curl $image -o \"" & downloadFilePath & "\"
echo \"Image downloaded\"
else
echo \"Image does not exist\"
fi
")
if shellOutput is "Image does not exist" then
copy imageURL to the end of missingImages
else
copy imageURL to the end of downloadedImages
end if
end repeat
set AppleScript's text item delimiters to delimitHolder
set missingImageDialogText to ""
if (count of missingImages) > 0 then
set missingImageDialogText to "The following images failed to download:"
repeat with anImage in missingImages
set missingImageDialogText to missingImageDialogText & return & anImage
end repeat
end if
display dialog "The script completed." & return & (count of downloadedImages) & " out of " & imageCount & " images were successfully downloaded to: " & saveFolder & return & missingImageDialogText buttons {"Cancel", "OK"} default button "OK"
Just to try to help troubleshoot⊠hereâs a version that still uses the base file path, but it also logs a plain text file of all the URLâs it was trying to load to the downloaded images folder.
use AppleScript version "2.4" -- Yosemite (10.10) or later
use scripting additions
set baseURL to "http://jpl.a.bigcontent.io/v1/static/"
set theSpreadsheet to choose file with prompt "Please select your csv file with the image URL's" of type {"csv"} without multiple selections allowed
tell application "Finder"
set downloadsFolder to path to downloads folder
set saveFolder to ((downloadsFolder as text) & "Script Downloaded Images:")
if not (exists saveFolder) then make new folder at downloadsFolder with properties {name:"Script Downloaded Images"}
set saveFolder to POSIX path of saveFolder
end tell
set spreadsheetData to read theSpreadsheet
set dataLines to the paragraphs of spreadsheetData
set theURLs to {}
set delimitHolder to AppleScript's text item delimiters
set dataForFile to ""
repeat with aLine in dataLines
set AppleScript's text item delimiters to ","
set newURL to baseURL & text item 1 of aLine
copy newURL to the end of theURLs
set dataForFile to dataForFile & newURL & return
end repeat
set dataSavePath to POSIX file (saveFolder & "URLs as Text.txt")
try
close access file dataSavePath
end try
set fileReference to open for access (dataSavePath) with write permission
set eof fileReference to 0
write dataForFile to fileReference
close access fileReference
set AppleScript's text item delimiters to delimitHolder
set imageCount to the count of theURLs
set missingImages to {}
set downloadedImages to {}
set delimitHolder to AppleScript's text item delimiters
repeat with imageURL in theURLs
set AppleScript's text item delimiters to "/"
set imageName to the last text item of imageURL
set downloadFilePath to saveFolder & imageName as text
set shellOutput to (do shell script "
image=\"" & (contents of imageURL as text) & "\"
response=$(curl --write-out %{http_code} --silent --output /dev/null $image)
if [ \"$response\" == \"200\" ]; then
curl $image -o \"" & downloadFilePath & "\"
echo \"Image downloaded\"
else
echo \"Image does not exist\"
fi
")
if shellOutput is "Image does not exist" then
copy imageURL to the end of missingImages
else
copy imageURL to the end of downloadedImages
end if
end repeat
set AppleScript's text item delimiters to delimitHolder
set missingImageDialogText to ""
if (count of missingImages) > 0 then
set missingImageDialogText to "The following images failed to download:"
repeat with anImage in missingImages
set missingImageDialogText to missingImageDialogText & return & anImage
end repeat
end if
display dialog "The script completed." & return & (count of downloadedImages) & " out of " & imageCount & " images were successfully downloaded to: " & saveFolder & return & missingImageDialogText buttons {"Cancel", "OK"} default button "OK"
Take a look and see if the URLâs look right in the text file. Maybe post part of the text file to the forum or PM me another dropbox link to what it saved so we can take a look, maybe itâll give me something to go on.
Another possible step is for me to simplify the CURL line and return terminal errors to Applescript.
Actually, come to think of itâŠ
Please open terminal and at the command prompt paste:
and hit enter and then post the response in Terminal here. Maybe there will be a CURL error that can help zoom us in.
Hi,
I ran the script on another computer and it worked fine
This is what terminal brought back on the machine that doesnât work
Last login: Tue May 2 09:30:55 on console
JDM10SHA00039:~ leo.eastland$ curl http://jpl.a.bigcontent.io/v1/static/bl_006806_a -o ~/Downloads/Script\ Downloaded\ Images/bl_006806_a
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:â:-- --:â:-- --:â:-- 0
Warning: Failed to create the file /Users/leo.eastland/Downloads/Script
Warning: Downloaded Images/bl_006806_a: No such file or directory
curl: (23) Failed writing received data to disk/application
and this is what terminal returned on the machine that the script works on
Last login: Fri Apr 21 18:46:10 on console
JDM11SHA00010:~ capture.station$ curl http://jpl.a.bigcontent.io/v1/static/bl_006806_a -o ~/Downloads/Script\ Downloaded\ Images/bl_006806_a
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:â:-- --:â:-- --:â:-- 0
0 3088k 0 0 0 0 0 0 --:â:-- --:â:-- --:â:-- 0Warning: Failed to create the file /Users/capture.station/Downloads/Script
Warning: Downloaded Images/bl_006806_a: No such file or directory
curl: (23) Failed writing body (0 != 11748)
Iâll cross reference the URL text files and see if I can see any anomalies.
Both these terminal responses are reporting failures. Did you have the âScript Downloaded Imagesâ folder that the script makes present in your Downloads folder on both machines when you ran the terminal command? Thatâs where I left it downloading to.
Alternately, I can just take that out of the path and download directly to your ~/Downloads folder:
Please try that.
Thanks,
Tom.
Sorry my bad tested again here you go
JDM11SHA00010:~ capture.station$ curl http://jpl.a.bigcontent.io/v1/static/bl_006806_a -o ~/Downloads/bl_006806_a
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 3088k 100 3088k 0 0 3519k 0 --:â:-- --:â:-- --:â:-- 3517k
JDM10SHA00039:~ leo.eastland$ curl http://jpl.a.bigcontent.io/v1/static/bl_006806_a -o ~/Downloads/bl_006806_a
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:â:-- --:â:-- --:â:-- 0
OK, well, got me. It worked in the top instance, didnât work in the bottom instance, and there was no error returned when it didnât work.
So weâve ruled out something weird about the Applescript that leads to the command being fed to terminal incorrectly (character encoding, escaping spaces, stuff like that.) Because a valid command fed directly into terminal still doesnât download the file.
Iâm afraid Iâm at a loss here for what could be going wrong on the machine where it doesnât work.
- Tom.
Yeah I am also at a loss with this, I though it maybe the version of curl but both machines operating exactly the same.