ARM64 Linux Support

[iamazeem]

ARM64 Linux Support

[liquidaty]

Great. merged "termcap" branch with main

[liquidaty]

Hm, it seems to work fine on a vanilla native ARM64 Linux setup; it is detecting termcap based on whether "include termcap.h" succeeds-- is it possible your cross-compile setup system has termcap.h in the include path but not in the lib? I suppose we could add a check for -ltermcap in addition to the include, but would prefer not to unless there is a legit scenario where -ltermcap could fail while #include <termcap.h> succeeds?

[liquidaty]

That said for reasons that may or may not be unrelated, I see tests failing on arm64 so we will look into that

[liquidaty]

above failed tests fixed (unrelated to this #14)

[iamazeem]

Yes, you are right. The termcap.h is in the path but for the host system i.e. x64 and not for ARM64. And, there is no libtermcap installed for the ARM64 cross-compiler.

[liquidaty]

Please check branch "termcap" and lmk if this issue is not resolved

[iamazeem]

Built locally for ARM64 from branch https://github.com/liquidaty/zsv/tree/termcap and the build was successful. 👍

Also, ran the binary with qemu and there were no error:

$ qemu-aarch64 -L /usr/aarch64-linux-gnu ./arm64-linux-gcc/bin/zsv version
zsv version v0.0.1-alpha-76-ge537972

$ qemu-aarch64 -L /usr/aarch64-linux-gnu ./arm64-linux-gcc/bin/zsv
zsv: streaming csv processor

Usage:
  zsv version: display version info (and if applicable, extension info)
  zsv (un)register [<extension_id>]: (un)register an extension
  zsv help [<command>]
  zsv license [<extension_id>]
  zsv <command> <options> <arguments>: run a command on data (see below for details)
  zsv <id>-<cmd> <options> <arguments>: invoke command 'cmd' of extension 'id'

Options common to all commands:
  -L,--limit-rows <n>: limit processing to the given number of rows (including any header row(s))
  -c,--max-column-count: set the maximum number of columns parsed per row. defaults to 1024
  -r,--max-row-size: set the minimum supported maximum row size. defaults to 64k
  -B,--buff-size: set internal buffer size. defaults to 256k
  -t,--tab-delim: set column delimiter to tab
  -O,--other-delim: set column delimiter to other column
  -q,--no-quote: turn off quote handling
  -v,--verbose: verbose output

Commands:
  select: extract rows/columns by name or position and perform other basic and 'cleanup' operations
  sql: run ad-hoc SQL
  count: print the number of rows
  desc: describe each column
  pretty: pretty print for console display
  flatten: flatten a table consisting of N groups of data, each with 1 or
           more rows in the table, into a table of N rows
  2json: convert to json
  2tsv : convert to tab-delimited text
  serialize: convert into 3-column format (id, column name, cell value)
  stack: stack tables vertically, aligning columns with common names
No extensions configured

(No extended commands)