Replies: 5 comments 6 replies
-
|
Sorry I cannot share this PDF file, but its not a totally healthy file: |
Beta Was this translation helpful? Give feedback.
-
|
somewhat surprisingly: The first few lines look like this: And there are many more Let me know if you need any other info from the file... |
Beta Was this translation helpful? Give feedback.
-
|
That's a lot of posts. There is no bug, and I think I can address what seems surprising at first sight. |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the detailed answers. All makes sense except 2.: I afraid I still don't get it. I don't see how the internal details matter: how can the existence of a temporary variable Its some kind of magic: when I look at "A" and "B" below, I assume the interpreter is free to change my code in A into B... Some fundamental assumption about programming is broken in my brain for this not be be so! A>> doc = fitz.open("f.pdf")
>> p = doc[0]
>> type(p)
fitz.fitz.Page
>> list(p.annots())
['FreeText' annotation on page 0 of f.pdf,
...
'Ink' annotation on page 0 of f.pdf,
'Ink' annotation on page 0 of f.pdf,
...
'FreeText' annotation on page 0 of f.pdf]
>> del p
>> del docB>> doc = fitz.open("f.pdf")
>> type(doc[0])
fitz.fitz.Page
>> list(doc[0].annots())
ValueError: orphaned object: parent is None |
Beta Was this translation helpful? Give feedback.
-
|
I also considered: CD |
Beta Was this translation helpful? Give feedback.
-
|
Let me make another try to explain what happens. Here is a session that may help: In [1]: import fitz
In [2]: doc = fitz.open("new-annots-0.pdf")
In [3]: annots = [] # list of annots on page 0
In [4]: for a in doc[0].annots(): # this is unproblematic
...: annots.append(a)
...:
In [5]: # This works!
In [6]: # But the list is mostly unusable, because the page is deleted now.
In [7]: # None of its items has a parent:
In [8]: len(annots) # no problem, as we do not touch items here
Out[8]: 15
In [9]: print(annots) # boom!
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
...
ValueError: orphaned object: parent is None
In [10]: tuple(annots) # boom!
Out[10]: ---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
...
ValueError: orphaned object: parent is None
In [11]: annots[0] # boom!
Out[11]: ---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
...
ValueError: orphaned object: parent is NoneSo it is not |
Beta Was this translation helpful? Give feedback.
-
|
I somewhat hesitate to muddy the waters further but the original error reported by one of my users is in fact with the msgs = []
# Any of these might indicate something more complicated than a scan
# and hence we should be safe and just render the page. We only try to
# extract the bitmap under very conservative circumstances. It is not
# safe to assume that if there is a single image on the current page
# then that is the scan - e.g., student annotates pdf using xournalpp and
# then stamps a smiley-face `.png` there."
if p.get_links():
msgs.append("Has links")
if list(p.annots()):
msgs.append("Has annotations")
if list(p.widgets()):
msgs.append("Has fillable forms")
if p.get_text("text"):
msgs.append("Has text")(here If there was more further detail about the RuntimeError, I don't have it. Its also possible that was an earlier version of PyMuPDF or MuPDF. |
Beta Was this translation helpful? Give feedback.
-
|
This is of course hard to tell without further context on version number and potentially example files. My initial reaction to the chosen approach to answer the question itself ("page empty?") is, that it uses sledge hammers to crack nuts:
if p.annot_xrefs():
msgs.append("Has annotations, links or widgets")
if p.get_text("text"):
msgs.append("Has text") |
Beta Was this translation helpful? Give feedback.
-
|
Thank you! I have gone with a minimal fix for now, but may try to the Replacing the "sledge hammer" is #815. |
Beta Was this translation helpful? Give feedback.
-
|
Maybe my formulation was a bit drastic ... 😉. What I was aiming at was that we need not try extracting annotation objects if all we want to know is whether the page has annotations at all. And especially we need not extract all of them all for that check: Note: |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I'm not sure what the expected behaviour is, but I'm getting a ValueError from
.annots():I don't get the error if I send types:
I also don't get the error if reference a page directly:
I can't think why
p.annots()would be different thandoc[0].annots()...Beta Was this translation helpful? Give feedback.
All reactions