How to bypass File::Slurp when it is required by another module?

vitoco has asked for the wisdom of the Perl Monks concerning the following question:

Hi! This seems to be quite repetitive but I still cannot figure out how to solve this...

I'm using Email::MIME::CreateHTML to build an email that contents an HTML message with a PNG image that I want to include in the email encoded as BASE64. Everything seems to work OK except for the fact that in the email, when displayed by the recipient's email client, the image is broken. I've decoded the BASE64 part of the multipart body and found that the encoded file had 1 byte less that the original image file. So I binary compared the original and decoded files and found that the missing byte was because read_file in File::Slurp always does a CR/LF into NL substitution with the asumption that it is a text file just because I'm running in a Windows machine ($buf =~ s/\015\012/\n/g if $is_win32 ;), and this binary image just happens to have that sequence of bytes in it's header.

Is there a way to replace the read_file method, without editing the modules or rewriting a whole resolver, or to force it to use binmode? Well, Slurp accepts some options like binmode, but this obvously needs CreateHTML to send that option to it, and I'm not sure if that works with small binary files.

Thanks!

Comment on How to bypass File::Slurp when it is required by another module? Select or Download Code

Replies are listed 'Best First'.
Re: How to bypass File::Slurp when it is required by another module? (monkeypatch) by beech (Parson) on Sep 17, 2018 at 06:21 UTC
Hi, Where is the call to read_file exactly? That is what you should fix/bugreport to use binmode You can override in your program by "monkeypatch" (an emergency measure) `sub File::Slurp::read_file { ... } or local *File::Slurp::read_file = sub { ... };` [download]	[reply] [d/l]
Re^2: How to bypass File::Slurp when it is required by another module? (monkeypatch) by marto (Cardinal) on Sep 17, 2018 at 06:33 UTC
Last time I looked the module was borked rt://83126.	[reply]
Re^3: How to bypass File::Slurp when it is required by another module? (monkeypatch) by vitoco (Hermit) on Sep 17, 2018 at 16:49 UTC
BTW, I've submitted this issue for Email::MIME::CreateHTML at rt://127130. I hope that the maintainer looks at this, as this modue was updated this year.	[reply]
Re^4: How to bypass File::Slurp when it is required by another module? (monkeypatch) by Anonymous Monk on Sep 17, 2018 at 22:24 UTC
Re^5: How to bypass File::Slurp when it is required by another module? (monkeypatch) by vitoco (Hermit) on Sep 18, 2018 at 07:50 UTC
Re^5: How to bypass File::Slurp when it is required by another module? (monkeypatch) by Anonymous Monk on Sep 18, 2018 at 09:25 UTC
Re^2: How to bypass File::Slurp when it is required by another module? (monkeypatch) by vitoco (Hermit) on Sep 17, 2018 at 15:22 UTC
The call is in line 36 of Email::MIME::CreateHTML::Resolver::Filesystem (permalink to source). I said I was not sure about `read_file` in File::Slurp to use options when the file size is less than 100K (permalink to source), but it seems to happen only on default reads from pseudofiles and when no options where provided. I'll try to "monkeypatch" one of these functions/methods...	[reply] [d/l]
Re: How to bypass File::Slurp when it is required by another module? by Aldebaran (Curate) on Sep 17, 2018 at 22:53 UTC
One solution is to re-write with Path::Tiny. It took me some time, but I'm happy that File::Slurp gets used, required by nothing going forward. `use Path::Tiny; # reading files $guts = $file->slurp; $guts = $file->slurp_utf8; @lines = $file->lines; @lines = $file->lines_utf8;` [download] They say not to re-invent the wheel, but I think you have to fix the flats. Uri Guttman might be the least pleasant person I've "met" in perl circles, and I wouldn't be surprised if he had interpersonal reasons for not maintaining the module that people know him for.	[reply] [d/l]
Re^2: How to bypass File::Slurp when it is required by another module? by beech (Parson) on Sep 18, 2018 at 00:36 UTC
Hi, slurp_raw for verbatim file :)	[reply]