Lots of memory leaks

[scamacho23]

🐛 Describe the bug

Liszt has some memleak issues:

• whenever scanf is used, valgrind reports that Liszt leaks some memory (this one may be a stdio issue)
• whenever wordfree is used, valgrind reports two memory leaks
• the commands -clar and -cln have a number of memleak errors, notably invalid read sizes and invalid frees
• the commands -ln and -lar have a number of memory leaks

💣 Steps to reproduce

Use the commands with valgrind to see what happens.

[Bwoltz]

Here's the results from Valgrind using the following command:
valgrind --leak-check=full ./lst -cln

==5067==
==5067== HEAP SUMMARY:
==5067== in use at exit: 151 bytes in 3 blocks
==5067== total heap usage: 26 allocs, 23 frees, 40,586 bytes allocated
==5067==
==5067== Searching for pointers to 3 not-freed blocks
==5067== Checked 69,960 bytes
==5067==
==5067== 34 bytes in 1 blocks are definitely lost in loss record 1 of 3
==5067== at 0x4C31D2F: realloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==5067== by 0x10AE33: print (in liszt/liszt/lst)
==5067== by 0x10AF53: cJSON_PrintUnformatted (in liszt/liszt/lst)
==5067== by 0x10EAA7: getCurrentNotePath (in liszt/liszt/lst)
==5067== by 0x10E966: getCurrentNote (in liszt/liszt/lst)
==5067== by 0x1116EB: clearNotes (in liszt/liszt/lst)
==5067== by 0x11288B: main (in liszt/liszt/lst)
==5067==
==5067== 117 (16 direct, 101 indirect) bytes in 1 blocks are definitely lost in loss record 3 of 3
==5067== at 0x4C31D2F: realloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==5067== by 0x4F49790: w_addword (wordexp.c:176)
==5067== by 0x4F49790: wordexp (wordexp.c:2425)
==5067== by 0x11180E: clearNotes (in liszt/liszt/lst)
==5067== by 0x11288B: main (in liszt/liszt/lst)
==5067==
==5067== LEAK SUMMARY:
==5067== definitely lost: 50 bytes in 2 blocks
==5067== indirectly lost: 101 bytes in 1 blocks
==5067== possibly lost: 0 bytes in 0 blocks
==5067== still reachable: 0 bytes in 0 blocks
==5067== suppressed: 0 bytes in 0 blocks
==5067==
==5067== ERROR SUMMARY: 2 errors from 2 contexts (suppressed: 0 from 0)
==5067== ERROR SUMMARY: 2 errors from 2 contexts (suppressed: 0 from 0)

"Loss record 1 of 3" seems to be referring to the getCurrentNodePath() function in helper.c

[scamacho23]

Yes! Thanks for sharing. I currently don't have the bandwidth to fix this, but I'd invite you to take wack at it because it's definitely an issue.

[Bwoltz]

Yea, I'll give it a shot

[Bwoltz]

Submitted a pull request

[scamacho23]

Thanks! Looks great. I only found those four commands to have leaks (although I admittedly checked through a while ago). Some others may have cropped up in the meantime (if you have the energy to look through -- once again, very very appreciated).

[cscortes]

This is still marked open, did you need help. I was thinking I could run some SAST tools to see if I could find anything? Let me know, otherwise maybe mark it as closed?