Posts tagged "url"

Acrobat/Reader 10: Save button not working for PDFs referenced from long URLs

Issue

If you are using Adobe Acrobat/Reader in Firefox/Chrome to view PDF files from long URLs you may notice that  the Save button is not working as expected.  It may be disabled (greyed out), or it may be enabled, but clicking on the button does not bring up the Save dialog to save the PDF to the local file system.

In some organisations the PDF files may be loaded from complex applications which require long URLs to serve parameters with the PDF.  For example:

http://helpx.adobe.com/content/dam/kb/en/837/cpsid_83709/attachments
/Acrobat_Enterprise_Administration.pdf?timestamp=t1115222011102&a=1234567890
123456789012345678901234567890123456789012345678901234567890123456789
&dkmtid=doc&ext=.PDF#FDF=http://server/XML.xfd

The URL above will show the PDF, but the Save button will not work as expected:

Reason

This is an issue with some versions of Adobe Acrobat/Reader 10 in Firefox/Chrome, and will be resolved in a later update for Acrobat/Reader.  The issue occurs when the URL contains more than 259 characters, as Firefox attempts to use this URL as the file-name in the Save dialog.  Such a long URL can exceed a file system limit and therefore the Save dialog does not appear.

Workarounds

  • Use shortened URLs to avoid the issue.
  • Use Acrobat/Reader in Internet Explorer.
  • Use Acrobat/Reader 9 with Firefox/Chrome.

Solution

This issue does not happen with the latest version of Acrobat/Reader 11, and will be fixed in a future release of Acrobat/Reader 10.

reference: (183411841/3319248)

VN:F [1.9.22_1171]
Was this helpful? Please rate the content.
Rating: 5.4/10 (14 votes cast)

CQ5.5: MCM emails contain URLs to author instead of publish

Issue

If you are using the Newsletter feature in MCM to send emails to lists of contacts you may notice that the URLs for assets in the email are pointing to the author instance, rather than the publish instance.  The links to the author instance will not work for external users, so the assets in the email will not show up.

Reason

This is a documentation issue as there is some further configuration required to substitute the links in the newsletter email so they point to the publish instance, rather than the author instance.  This will be addressed in a later version of CQ.

Solution

To resolve this issue in the meantime you should do the following steps:

  1. login to the system console on author (e.g. http://localhost:4502/system/console/configMgr)
  2. in the configuration tab search for the “Day CQ MCM Newsletter” bundle and open it
  3. take note of the fake host name in “Lookup for Sender Host” field (by default it is –publish–), i.e. the field should have http://–publish– in it
  4. open CRXDE Lite and goto /etc/map/http
  5. create a Node named –publish– of type sling:Mapping
  6. add a String property sling:match with value –publish–.80
  7. add a String property sling:redirect with value of your publish instance host name or IP including port(e.g. http://publishhost:4503)
  8. click Save All

Now when you send the newsletter it should substitute the author URL with the publish URL.

reference: (37433/CQ5-15440)

VN:F [1.9.22_1171]
Was this helpful? Please rate the content.
Rating: 8.5/10 (4 votes cast)

Acrobat/Reader: Attach to Email button not working for PDFs referenced from long URLs

Issue

If you are using servlets or scripts to serve your PDF files in the browser, and you then want to use the standard “Attach to Email” button on the Acrobat/Reader toolbar (also available under the File menu) to forward the PDF to another user by email, you may notice that the PDF does not get attached to the email in some situations.

Reason

When you click on the Acrobat/Reader “Attach to Email” button on the toolbar it sends a request to your default email client to compose a new message.  It will also attempt to create the PDF attachment using the file name specified in the URL where the actual PDF is stored.

The URL may be too long to create a PDF file as attachment if it contains very long parameters as follows:

http://localhost:8080/Simple_Servlet/DMShowPDF?command=getdocuments&PDF=path/123_4567890abcdefghijklmnopqrstuvwxyz_0_0.pdf&sig=0123456789abcdefghijklmnopqrstuv&ts=0123456789012&category=new&stage=EDITING%3B%3B%3B1%3B%3B%3B&docid=S0123456789abcdefghijklmnopqrst&ext=.PDF

In this case the filename of the resulting PDF attachment would be:

123_4567890abcdefghijklmnopqrstuvwxyz_0_0.pdf&sig=0123456789abcdefghijklmnopqrstuv&ts=0123456789012&category=new&stage=EDITING%3B%3B%3B1%3B%3B%3B&docid=S0123456789abcdefghijklmnopqrst&ext=.PDF

This can exceed operating system limits, and therefore the attachment does not get created/added to the email message.

Workaround

Simply reducing the length of the URL, particularly the string which will be used for the PDF attachment file name, will allow the PDF attachment to be created.  Even just reducing it by a few characters is sufficient.

Solution

A cleaner solution however, would be to re-structure the URLs so they are shorter, and/or move the PDF parameter to the last parameter position as follows:

http://localhost:8080/Simple_Servlet/DMShowPDF?command=getdocuments&sig=0123456789abcdefghijklmnopqrstuv&ts=0123456789012&category=new&stage=EDITING%3B%3B%3B1%3B%3B%3B&docid=S0123456789abcdefghijklmnopqrst&ext=.PDF&PDF=path/123_4567890abcdefghijklmnopqrstuvwxyz_0_0.pdf

In this case, the attachment file name will be:

123_4567890abcdefghijklmnopqrstuvwxyz_0_0.pdf

reference: (182987737)

VN:F [1.9.22_1171]
Was this helpful? Please rate the content.
Rating: 3.3/10 (12 votes cast)

LiveCycle ES: old XDP templates with invalid WSDL URLs used for running processes

Issue

 If you have migrated from LC7 to LC ES and you were using webservices in your XDP templates in LC7, you may have modified these templates after the mgiration to LC ES to use updated WSDL URLs.  You may notice however that even for process instances started in LC ES the old template version is still being used, and so the invalid WSDL URL which means the tasks will no longer work as expected.

Reason

In LC7 we set both the template_url and hardened_template_url fields in the tb_form_data table to reference the forms used in LC7.  In LC ES we no longer use the hardened_template_url field for new LC ES processes.  However for processes migrated from LC7, we still use this field to reference the exact version of the form that was used to start the process.

The template_url field in tb_form_data contains only the URL to the form in the repository (/Forms/form1.xdp).  The hardened_template_url contains a timestamp with the URL to reference a particular version of the form in the repository, and can therefore referenence different versions of the form (@1271757648748/Forms/from1.xdp).

Solution

If you set the contents of hardened_template_url to empty in tb_form_data, then LC ES will use the template_url by default, which does not have a timestamp, and therefore will always refer to the latest version of the XDP form in the repository.  This will ensure that your updated XDP template with updated WSDL URL will be used for each process.

You can run the following SQL statements first to see how many rows are going to be changed:

Statement1:

This is used to check which forms are being used for active tasks which have a hardened URL set.  These are the forms that are going to be modified by the SQL query, so they should relate to the forms you have modified for the WSDL changes.

SELECT DISTINCT f.template_url, f.hardened_template_url
FROM tb_form_data f
WHERE f.task_id IN (SELECT t.id
 FROM tb_task t
 WHERE t.status < 100)

Statement2:

This is used to ensure that there are no NULL entries in the template_url field.  The template_url field is the field we are going to use to refer to the modified form. Therefore, this statement should not return any rows.  You should probably run this statement again checking for “ “ instead of NULL values just to be sure there are no empty strings in template_url.

SELECT t.form_data_id, f.id, f.template_url, f.hardened_template_url
FROM tb_task t, tb_form_data f
WHERE (f.template_url IS NULL) AND
 (f.hardened_template_url IS NOT NULL) AND
 f.task_id IN (SELECT t.id
 FROM tb_task t
 WHERE t.status < 100)

Statement 3:

If statement2 does not return any rows, then you can run this SQL to actually make the required changes in hardened_template_url.

UPDATE tb_form_data
SET hardened_template_url = NULL
WHERE (template_url IS NOT NULL) AND
 (hardened_template_url IS NOT NULL) AND
 task_id IN (SELECT t.id
 FROM tb_task t
 WHERE t.status < 100)

Notes

If you are unsure about applying these SQL queries to the Database, or if this occuring in a production environment, then contact your support representative to discuss the workaround and to verfiy the results of the first two SQL statements.

reference: (181544186)

VN:F [1.9.22_1171]
Was this helpful? Please rate the content.
Rating: 0.0/10 (0 votes cast)

LiveCycle Workbench: which URLs are used for communication with LC server

Information

 If you are using a proxy between Workbench and the LiveCycle ES Server, you may want to limit the connection to provide a higher level of security.  For the standard installation Workbench communcates with the LiveCycle server over HTTP with servername and port number as follows:

http://<server>:<port>

Using a proxy server, you could limit the connection further to just two URLs:

http://<server>:<port>/DocumentManager
http://<server>:<port>/soap/sdk

This is subject to change, especially in later versions and must be tested thoroughly to ensure there are no conflicts or unexpected problems.  Adobe will investigate issues related to such a configuration but in some cases the only solution may be to return to the default settings, without limiting the connection to specific URLs.

reference: (181300310)

VN:F [1.9.22_1171]
Was this helpful? Please rate the content.
Rating: 0.0/10 (0 votes cast)