Request feature: supporting larger output files

[jywu20]

Is your feature request related to a problem? Please describe.
When I try to parse a bands.out file containing about 10000 lines, I get this error:

ERROR: PCRE.exec error: JIT stack limit reached
Stacktrace:
 [1] error(s::String)
   @ Base .\error.jl:33
 [2] exec(re::Ptr{Nothing}, subject::String, offset::Int64, options::UInt32, match_data::Ptr{Nothing})
   @ Base.PCRE .\pcre.jl:197
 [3] exec_r_data
   @ .\pcre.jl:210 [inlined]
 [4] match(re::Regex, str::String, idx::Int64, add_opts::UInt32)
   @ Base .\regex.jl:371
 [5] match
   @ .\regex.jl:369 [inlined]
 [6] match
   @ .\regex.jl:388 [inlined]

The error is thrown at this line.

Describe the solution you'd like
Appropriate pre-processing to avoid blowing up the regex module, or maybe increasing the JIT stack as is suggested by the error message?

[singularitti]

Hi @jywu20, I have never thought 10000 lines of data would make it out of memory. I am not familiar with how PCRE (the regular expression engine Julia uses) works but it seems that it is a problem of PCRE and not unique to our code or Julia:

1. pcre error -27 (JIT stack limit) on long regex string
2. PCRE JIT stack size limit
3. RegEx not working for long pattern PCRE's JIT compiler stack limit - PHP7

[jywu20]

Yeah, I have checked all the three links before I submitted this issue. The location of this error is quite straightforward: it has nothing to do with the logic of the QuantumESPRESSO part. What concerns me is there seems to be no well-documented method to increase the memory allocated for PCRE. This post provides a ccall code to increase the allocation, which I tried but it didn't work.

[github-actions]

Stale issue message