fix mkshort unique

[tsteven4]

1. we were off by one when deciding if we could just concatenate
   the generated suffix to the name, failing to use the last available
   character.
2. when attempting to copy the generated suffix to a presumably
   shortened name, and the generated suffix was longer than the name,
   the target of the copy was outside the name buffer. In the common
   case that the rank of size_t was greater than the rank of int, the
   target was well beyond the end of name, not before as one would expect
   with signed arithmetic.

With the new algorithm when the target length is insufficient to
fit both the name and the suffix we will only truncate the name as
required to fit the truncated name and the complete suffix in the
target length(as opposed to the original length of the name).
We fatal if this is not possible.

Add a test case to exercise the make unique code.
@robertlipe please verify that the generated references mkshort6.csv, mkshort5.csv, mkshort4.csv have generated names as desired.

[GPSBabelDeveloper]

LGTM Thanx! Instead of this suffering (too late) why not just make the the lenght unsigned? It's never sensible for it to be negative ... or even really smaller than, oh, "six" or so in practice. I do agree with your sense of increased fatalism in the code. There are some cases, mostly pathological, that just plain won't fit. I also later developed a bad feeling that the code was very English-oriented, but the feelign wasn't bad enough to include tables of words of the digits in the top 45 languages. We should also assert that '.' is never in the blacklist. I do also, unsurprisignly, agree with your (heroic) efforts to preserve fidelity for those cases. Those seem spot-on. For a different spin on all this, peek at shortname.cc , which was never finished but was my hobby while in flight during the work commute for a trip or two, before we were even C++, I think. About half the body was this terrible "test driven design" approach of a table of 2003 (ish) geocache names and how they'd be reduced to receivers of that era. Then again, as our own mkshort.cc has become more QString-centric, it's also become less terrible to look at on its own. Trivia: in all my geocaching escapades, there were only two cases where I know there were geocaches I couldn't find because they were lost by GPSBabel on the way to the receiver. One involved very long geocache names with a sequence number that wasn't at the end - we collapsed those all into one and most of the series just wasn't on my hardware. The second I only caught because I had two _different_ Garmins in my hunting at that time. An often missed (and the kind of thing you'd approximately never find via reverse engineering) nuance of the Garmin USB protocol is that on certain models, if the sent packet is exactly 256 characters (255?) you must insert a following packet that's basically empty to "push" that one on through, otherwise it will be lost. So I'd used a shell script that loaded one GPX file (maybe a merge of a couple...) to my StreetPilot and my GPSMap 60Cx (60CSx?) and one receiver had a specific point and the other didn't. Took a serious technology dive to find the truth on that one.

…

On Thu, Aug 31, 2023 at 8:12 AM tsteven4 ***@***.***> wrote: 1. we were off by one when deciding if we could just concatenate the generated suffix to the name, failing to use the last available character. 2. when attempting to copy the generated suffix to a presumably shortened name, and the generated suffix was longer than the name, the target of the copy was outside the name buffer. In the common case that the rank of size_t was greater than the rank of int, the target was well beyond the end of name, not before as one would expect with signed arithmetic. With the new algorithm when the target length is insufficient to fit both the name and the suffix we will only truncate the name as required to fit the truncated name and the complete suffix in the target length(as opposed to the original length of the name). We fatal if this is not possible. Add a test case to exercise the make unique code. @robertlipe <https://github.com/robertlipe> please verify that the generated references mkshort6.csv, mkshort5.csv, mkshort4.csv have generated names as desired. ------------------------------ You can view, comment on, or merge this pull request online at: #1168 Commit Summary - e286c7b <e286c7b> fix mkshort unique. - 9cb23ee <9cb23ee> fix reference mode. - 122a735 <122a735> add notes for future enhancements - d936a94 <d936a94> fix testcase cut and paste booboo File Changes (8 files <https://github.com/GPSBabel/gpsbabel/pull/1168/files>) - *M* mkshort.cc <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-f52a7a9b87deb261fda2847180d2f4d7423753bbf22de3a8af530c33ddb7b272> (49) - *A* reference/mkshort.csv <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-caea3713b8874e5d24ff078e98159831849eb564317f4b3c553fdcbac2c461b8> (151) - *A* reference/mkshort.style <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-24fa103798ffb96cc6d5d0edb136163a6b344d650d05f586d1121f081b4a0c86> (24) - *A* reference/mkshort3.log <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-222564face2e1428a82ad1aca9b9dd75ce427a72e3a1ead5d9bee070b9eb4017> (1) - *A* reference/mkshort4.csv <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-16587ea7d0414576df19b2e01353965a2816189771cb9a8ac796c9b0c1d0c4e0> (150) - *A* reference/mkshort5.csv <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-f3a0b6d130070096f46cbb6e8122b5713ed13b82aa6c4fa62e37aba197a7dc04> (150) - *A* reference/mkshort6.csv <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-4a9931921ab1fd3f2a99f996612219d74f7dddc38feb50e02983c49d66698705> (150) - *A* testo.d/mkshort.test <https://github.com/GPSBabel/gpsbabel/pull/1168/files#diff-869f0f3261fa0412ce1dcefa630d9025c7fe00a023fdae97172f3ae31637e926> (13) Patch Links: - https://github.com/GPSBabel/gpsbabel/pull/1168.patch - https://github.com/GPSBabel/gpsbabel/pull/1168.diff — Reply to this email directly, view it on GitHub <#1168>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC3VADZSGFQBOSZ65JNW45DXYCEVNANCNFSM6AAAAAA4GB5KMI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>