mirror of
https://github.com/tonydamage/nux-env.git
synced 2025-12-11 13:24:28 +01:00
109 lines
3.1 KiB
Bash
109 lines
3.1 KiB
Bash
## #nux.json NUX Script JSON Helper library
|
|
##
|
|
## #Public functions
|
|
@namespace nux.json. {
|
|
## nux.json.start
|
|
## Starts operation on new JSON. Resets write operations stack and JSON
|
|
## source file to empty (useful for creating JSON from scratch.)
|
|
|
|
function :start {
|
|
nux_json_opstack=".";
|
|
nux_json_source="-n";
|
|
}
|
|
|
|
## nux.json.open <file>
|
|
## Opens JSON File for reading and writing using *nux.json.open* and
|
|
## *nux.json.write* commands.
|
|
##
|
|
## In order to work with empty JSON use *-n* as *file*. This allows
|
|
## for creating new JSON from scratch.
|
|
##
|
|
## NOTE: Open does not reset operation stack. To reset operation stack
|
|
## and opened file use *nux.json.start*
|
|
function :open file {
|
|
if [ -f "$file" ]; then
|
|
nux_json_source="$file"
|
|
fi
|
|
}
|
|
|
|
## nux.json.read <path>
|
|
## Reads *path* from currently opened JSON file.
|
|
## NOTE: Read does not see changes performed by *nux.json.write* unless
|
|
## these changes were flushed using *nux.json.flush*.
|
|
function :read path {
|
|
jq -r ".$path" "$nux_json_source";
|
|
}
|
|
|
|
## nux.json.write <path> <value>
|
|
## Adds write operation to action stack. Writes are not performed
|
|
## immediately, but rather when *nux.json.flush* is invoked.
|
|
## This allows for batching of operations or opting out of actually
|
|
## modifying file.
|
|
function :write path value {
|
|
nux_json_opstack="${nux_json_opstack} | .$path |= \"$value\""
|
|
}
|
|
|
|
function :write.raw path value {
|
|
nux_json_opstack="${nux_json_opstack} | .$path |= $value"
|
|
}
|
|
|
|
## nux.json.flush [<target>]
|
|
## Flushes any write operations to specified *target* file.
|
|
## If *target* is not specified JSON is outputted to *STDIN*.
|
|
##
|
|
## NOTE: Flush does not reset operation stack. To reset
|
|
## operation stack and opened file use *nux.json.start*
|
|
##
|
|
function :flush target {
|
|
if [ -n "$target" ]; then
|
|
local write_target="$target";
|
|
if [ "$nux_json_source" == "$target" ]; then
|
|
write_target=$(mktemp "$(dirname "$target")/tempXXXXXX.json")
|
|
fi
|
|
jq -r "$nux_json_opstack" "$nux_json_source" > "$write_target"
|
|
if [ "$nux_json_source" == "$target" ]; then
|
|
mv -f "$write_target" "$target"
|
|
fi
|
|
else
|
|
jq -r "$nux_json_opstack" "$nux_json_source"
|
|
fi
|
|
}
|
|
|
|
## nux.json.shorthands
|
|
## Exposes shorthands for writing and reading from JSON file.
|
|
## *njw* - nux.json.write
|
|
## *njr* - nux.json.read
|
|
##
|
|
function :shorthands {
|
|
function njw {
|
|
nux.json.write "$@"
|
|
}
|
|
function njr {
|
|
nux.json.read "$@"
|
|
}
|
|
}
|
|
|
|
## # Usage Notes
|
|
##
|
|
## ## Operation stack and flush
|
|
##
|
|
## As mentioned in documentation for *nux.json.flush* write stack
|
|
## is not removed, but rather kept alone, separatly from reference
|
|
## to open file. This allows for having modification template
|
|
## which could be executed on multiple files.
|
|
##
|
|
## The following example adds meta.author and meta.email
|
|
## to every JSON in directory.
|
|
##
|
|
## *nux.json.start*
|
|
## *nux.json.write* meta.author "Tony Tkacik"
|
|
## *nux.json.write* meta.email "example@example.com"
|
|
## *for* f *in* "*.json";
|
|
## *do*
|
|
## *nux.json.open* "$f"
|
|
## *nux.json.flush* "$f"
|
|
## *done*;
|
|
##
|
|
##
|
|
|
|
}
|