proposal on improving uudecoding

proposal on improving uudecoding

[Vladimir Volovich]

Hi,

Some mailers insert blank lines into a uuencoded files,
so that they look like the following:

MUNH]N<SL0?6K\ZOQ0@+_6[_#!:<D&J7J5/9OOJ_U&:]QB_O5IF-J]F7P'V[A
MJM8C^QZ$WW?V0CK=N@O!TN5YES/]=<F)`W8<6JA-#8@!CJBQR!SPW)SD4%^^
M`#8G';;F1G+/#O"64@OP9%">)T^T1!_HR=-`SX5#@"6;-_F87XJW8"O0_".=

M>6,2]B*E.K$AKQY1_QAS3%621_2B9/]P*;]ES["?*?W;([22UEC*J!1L\F+]

M6I&`;=Q9PN<0UA_WUOFL:/&1'2SHD]<[Q8M[9=E@]2;BEXL`ZLF..VE_:<SU

M>Q%N=(S6<^D>;A61,@>LK#35S\O>01`(%`F'N+^Z#A\$*4?VI5$UX^$U$_9!

M0;OE<\(J;F#0L:V$L?WH^EB49=>L7@>6'_:7&-]*(K"LTE;@H^\8&:?%=MT5
MR`SB;]7;_3$=UG'ZBP0V=_X@IPW.'U]Q*83/0:-B^+DNP-QL0QZ5F<\#3H[P

Gnus incorrectly uudecodes such articles because of these blank lines.
Would it be possible to skip blank lines (not beginning with /^M/
regexp) before uudecoding? This will help a lot, especially when
decoding files splitted in several messages.

-- 
With best regards,
                   Vladimir.

Re: proposal on improving uudecoding

[Vladimir Volovich]

"SSR" == Stainless Steel Rat writes:

 VV> Some mailers insert blank lines into a uuencoded files, so that
 VV> they look like the following:

 SSR> I have yet to see one... but one that does is 100% broken.

Certainly. Hmm, actually, not 100% because these messages *can* be
restored in their original form by removing empty lines before
uudecoding. ;-)

 SSR> An MUA or MTA should never arbitrarilly change the contents of a
 SSR> message.

Yes, but such a weird MUAs *do* exist (mainly these are some FidoNet
MUAs which change the contents of included files because they
interpret the `>' characters as a citation prefixes and want to make
the included text more `readable' %-) ). And there *is* a simple
work-around: just remove blank lines automagically before uudecoding
(anyway, these blank lines should not appear inside a uuencoded texts,
and thus this change won't break anything).

-- 
With best regards,
                   Vladimir.

Re: proposal on improving uudecoding

[Lars Magne Ingebrigtsen]

Vladimir Volovich <vvv@vvv.vsu.ru> writes:

> Gnus incorrectly uudecodes such articles because of these blank lines.
> Would it be possible to skip blank lines (not beginning with /^M/
> regexp) before uudecoding? This will help a lot, especially when
> decoding files splitted in several messages.

I won't be adding code to handle this, but I've now added
`gnus-uu-pre-uudecode-hook', which can be used to fix broken uuencoded
text.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen