Durin should check for and report missing data files rather than crash

[KayDiederichs]

At the recent CCP4 @ DLS workshop, a student restored data (collected in June 2025 at I04) from tape but for some reason the restore operation was not complete, or we started to process the data before it finished. Anyway, out of the two data files xtal_18_2_000001.h5 and xtal_18_2_000002.h5 , only the first was available. According to the xtal_18_2_master.h5, the dataset has 1200 frames.

All my attempts to process the dataset with the Durin plugin (no matter whether copied from autoPROC or using the one from Github) failed with not-very-helpful messages:

master_file=/Users/dikay/projects/Kriti.Maitre/xtal_18_2_master.h5
[generic_data_plugin] - INFO - generic_open
+ library = </usr/local/autoPROC/autoPROC/bin/darwin/plugins-arm64/durin-plugin.so>
+ template_name = </Users/dikay/projects/Kriti.Maitre/xtal_18_2_master.h5>
+ dll_filename = </usr/local/autoPROC/autoPROC/bin/darwin/plugins-arm64/durin-plugin.so>
+ image_data_filename = </Users/dikay/projects/Kriti.Maitre/xtal_18_2_master.h5>
Durin plugin error:
src/plugin.c - line 304 in plugin_open:
Could not open /Users/dikay/projects/Kriti.Maitre/xtal_18_2_master.h5
H5F.c - line 412 in H5Fopen
H5Fint.c - line 1700 in H5F_open
H5FD.c - line 722 in H5FD_open
H5FDsec2.c - line 351 in H5FD__sec2_open:
unable to open file: name = '/Users/dikay/projects/Kriti.Maitre/xtal_18_2_master.h5', errno = 2, error message = 'No such file or directory', flags = 0, o_flags = 0
generic_getfrm: could not open /usr/local/autoPROC/autoPROC/bin/darwin/plugins-arm64/durin-plugin.so ier= -4
[generic_data_plugin] - INFO - generic_get_header
Program received signal SIGSEGV: Segmentation fault - invalid memory reference.
Backtrace for this error:
#0 0x1002a9037
#1 0x1002a83f3
#2 0x19ad5d6a3
#3 0x100ae34eb
#4 0x1001dcfd7
#5 0x1002a7bbb
#6 0x10020f4a7
#7 0x10028fd87
#8 0x1001e4ce7
#9 0x1002a82c7
#10 0x1002ddc6f

This should report, in my understanding, the non-availability of xtal_18_2_000002.h5 but instead it blames xtal_18_2_master.h5 .

To reproduce: a) David Waterman has reproduced this with the data at /dls/mx/data/mx43974/mx43974-5/processing/incoming/KD/Kriti.Maitre-crash . However, I believe that the problem of a missing / misleading error message is a general one.

[KayDiederichs]

Going beyond the issue of the lack of proper error message and crash, I wish I could at least process those data that I have, namely the 1000 frames in xtal_18_2_000001.h5 . IOW, in case of missing data files, make it possible to process the existing ones.