change the sed() parser to split out the command (aisdxX) from the args

change the parseQuotedArgs() to return a tuple (pattern, filename)
This commit is contained in:
2021-04-10 08:49:33 -04:00
parent ddbea490ea
commit eb68b76c8a
2 changed files with 22 additions and 20 deletions

View File

@@ -169,7 +169,7 @@ If the single-letter command is preceded by a number or number-range, then the e
The `i/a/d` commands should be preceeded by a line number, or range; `sed()` will *insert*, *append* or *delete* once for each line in the range. The `i/a/d` commands should be preceeded by a line number, or range; `sed()` will *insert*, *append* or *delete* once for each line in the range.
The ``x/X` patterns are wrapped in a pair of delimiter characters, typically /, although any other character is allowed (except space or any of `\^$()[]`). Valid X commands are: The ``x/X` patterns are wrapped in a pair of delimiter characters, typically `/`, although almost any other character is allowed other than space or one of `\^$()[]`. Valid X commands are:
``` ```
x/abcd/ x/abcd/
@@ -177,7 +177,7 @@ x/abcd/
x!ratio x/y! x!ratio x/y!
``` ```
Similarly, the s patterns are wrapped in a triplet of delimiter characters, typcially / also. If the search pattern has `()` groups, the replace pattern can refer to them with ``\1 \2`,etc. Valid 's' commands are Similarly, the s patterns are wrapped in a triplet of delimiter characters, typcially / also. If the search pattern has `()` groups, the replace pattern can refer to them with ``\1 \2`,etc. This `sed()` assumed the `g` suffix, it replaces multiple occurrences on a line. Valid 's' commands are
``` ```
s/toronto/Toronto/ s/toronto/Toronto/
@@ -267,6 +267,8 @@ In its present form, the module has these limitations:
* with sed, lines are parsed and saved one-line-at-a-time, so pattern matching to \n and \r does not work; sed cannot work over line boundaries * with sed, lines are parsed and saved one-line-at-a-time, so pattern matching to \n and \r does not work; sed cannot work over line boundaries
* this simple shell is different than [mpfshell](https://github.com/wendlers/mpfshell) in that this shell runs entirely on the target device. There is no allowance in this shell for transferring files in/out of the target. * this simple shell is different than [mpfshell](https://github.com/wendlers/mpfshell) in that this shell runs entirely on the target device. There is no allowance in this shell for transferring files in/out of the target.
* after a restart of your *MicroPython* board, you can invoke the shell with `import tf`; if you `^C` out of the shell, the second invocation of `tf` will have to be `import tf` followed by `tf.main()`, since the python interpreter caches the module and only loads it once per restart; you can intentionally restart the REPL prompt by hitting `^D` * after a restart of your *MicroPython* board, you can invoke the shell with `import tf`; if you `^C` out of the shell, the second invocation of `tf` will have to be `import tf` followed by `tf.main()`, since the python interpreter caches the module and only loads it once per restart; you can intentionally restart the REPL prompt by hitting `^D`
* for the `sed()` function and command line, the
[search](https://docs.micropython.org/en/latest/library/ure.html) pattern can have wildcards like ``\s`, `\w` and `\d`. The replace pattern cannot have *any* of these, and can only have `\0`, `\1`, etc
## Examples ## Examples

36
tf.py
View File

@@ -43,10 +43,10 @@ def grep(filename, pattern, numbers=False):
def sed(filename, sed_cmd, bak_ext=".bak"): def sed(filename, sed_cmd, bak_ext=".bak"):
# parse the sed_cmd # parse the sed_cmd
# group 1,3 are the n-start, n-end group 4 is command: aidsxX # group 1,3 are the n-start, n-end group 4 is command: aidsxX
a=re.search("^(\d*)([,-](\d+|\$))?\s*([sdaixX].*)",sed_cmd) a=re.search("^(\d*)([,-](\d+|\$))?\s*([sdaixX])(.*)",sed_cmd)
if not a: if not a:
raise ValueError("sed() failed; pattern must be a number-range followed by one of sdaixX; no changes applied") raise ValueError("sed() failed; pattern must be a number-range followed by one of sdaixX; no changes applied")
cmd=a.group(4) op,args=a.group(4),a.group(5)
s,e=1,1000000 s,e=1,1000000
if a.group(1): if a.group(1):
@@ -54,26 +54,25 @@ def sed(filename, sed_cmd, bak_ext=".bak"):
if a.group(3): if a.group(3):
e=1000000 if a.group(3)=='$' else int(a.group(3)) e=1000000 if a.group(3)=='$' else int(a.group(3))
op=cmd[0]
if op in "aid" and e-s==1000000: if op in "aid" and e-s==1000000:
raise ValueError("sed(a/i/d) should have a line number") raise ValueError("sed(a/i/d) should have a line number")
#print("sed command parser of <{}> returned {} {} {}".format(op,cmd,a.group(1),a.group(3))) #print("sed command parser of <{}> returned {} {} {}".format(op,cmd,a.group(1),a.group(3)))
if op in "sxX": if op in "sxX":
if len(cmd)<2 or cmd[1] in "\^$()[]": if len(args)<2 or args[0] in "\^$()[]":
raise ValueError("invalid sed argument: {}".format(cmd)) raise ValueError("invalid sed argument: "+op+args)
dl=cmd[1] dl=args[0]
if op=='s': if op=='s':
gs=re.search("s"+dl+"([^"+dl+"]*)"+dl+"([^"+dl+"]*)"+dl,cmd) gs=re.search(dl+"([^"+dl+"]*)"+dl+"([^"+dl+"]*)"+dl,args)
else: else:
gs=re.search("[xX]"+dl+"([^"+dl+"]*)"+dl,cmd) gs=re.search(dl+"([^"+dl+"]*)"+dl,args)
if not gs: if not gs:
raise ValueError("invalid sed search pattern: {}".format(cmd)) raise ValueError("invalid sed search pattern: "+op+args)
ss=gs.group(1) ss=gs.group(1)
if op=='s': if op=='s':
r=gs.group(2) r=gs.group(2)
sp=re.compile(ss) sp=re.compile(ss)
extra=a.group(4)[1:] + '\n' args=args + '\n'
try: try:
os.rename(filename,filename+bak_ext) os.rename(filename,filename+bak_ext)
@@ -95,7 +94,7 @@ def sed(filename, sed_cmd, bak_ext=".bak"):
continue # delete line continue # delete line
if op=='i' and m: if op=='i' and m:
#print("insert a line before {} <{}>".format(i,extra)) #print("insert a line before {} <{}>".format(i,extra))
g.write(extra) g.write(args)
h+=1 h+=1
if op in "aids": if op in "aids":
g.write(lin) g.write(lin)
@@ -104,7 +103,7 @@ def sed(filename, sed_cmd, bak_ext=".bak"):
h+=1 h+=1
if op=='a' and m: if op=='a' and m:
#print("append a line after {} <{}>".format(i,extra)) #print("append a line after {} <{}>".format(i,extra))
g.write(extra) g.write(args)
h+=1 h+=1
#f.write("--file modifed by sed()--\n") #f.write("--file modifed by sed()--\n")
return (i, h) return (i, h)
@@ -139,15 +138,16 @@ def _help():
ext_cmd('help') ext_cmd('help')
def parseQuotedArgs(st): def parseQuotedArgs(st):
#returns (pattern,file)
st=st.strip() st=st.strip()
if st[0]=="'": if st[0]=="'":
p=re.search("'(.*?[^\\\\])'",st) p=re.search("'(.*?[^\\\\])'\s*(\S*)",st)
if not p: if not p:
print("error in quoted pattern:",st) print("error in quoted pattern:",st)
return return
return p.group(1) return p.group(1),p.group(2)
else: else:
return st.split()[0] return st.split(None,1)
def main(): def main():
print("Simple shell: cp/copy mv/move rm/del cat/list cd dir/ls mkdir rmdir grep sed help") print("Simple shell: cp/copy mv/move rm/del cat/list cd dir/ls mkdir rmdir grep sed help")
@@ -183,9 +183,9 @@ def main():
continue continue
try: try:
if op=='grep': if op=='grep':
grep(rp[-1],p,numbers=True) grep(p[1],p[0],numbers=True)
else: else:
r=sed(rp[-1],p) r=sed(p[1],p[0])
if r: if r:
print("Lines processed: {} Lines modifed: {}".format(*r)) print("Lines processed: {} Lines modifed: {}".format(*r))
except (ValueError, OSError) as e: except (ValueError, OSError) as e:
@@ -216,7 +216,7 @@ def main():
except IndexError: except IndexError:
print("not enough argments; check syntax with 'help'") print("not enough argments; check syntax with 'help'")
except OSError: except OSError:
print("file/folder not found or cannot be writtencd") print("file/folder not found or cannot be written")
gc.collect() gc.collect()
if __name__=="tf": if __name__=="tf":