Magic command to compare files¶

Links: notebook, html, PDF, python, slides, GitHub

Some ways to display differences between files.

from jyquickhelper import add_notebook_menu
add_notebook_menu()

Two functions slighly different ¶

f1 = '''
def edit_distance_string(s1, s2):
    """
    Computes the edit distance between strings *s1* and *s2*.

    :param s1: first string
    :param s2: second string
    :return: dist, list of tuples of aligned characters
    """
    n1 = len(s1) + 1
    n2 = len(s2) + 1
    dist = numpy.full((n1, n2), n1 * n2, dtype=numpy.float64)
    pred = numpy.full(dist.shape, 0, dtype=numpy.int32)

    for j in range(1, n2):
        dist[0, j] = j
        pred[0, j] = 2
    for i in range(0, n1):
        dist[i, 0] = i
        pred[i, 0] = 1
    pred[0, 0] = -1

    for j in range(1, n2):
        for i in range(1, n1):
            c = dist[i, j]

            p = 0
            if dist[i - 1, j] + 1 < c:
                c = dist[i - 1, j] + 1
                p = 1
            if dist[i, j - 1] + 1 < c:
                c = dist[i, j - 1] + 1
                p = 2
            d = 0 if s1[i - 1] == s2[j - 1] else 1
            if dist[i - 1, j - 1] + d < c:
                c = dist[i - 1, j - 1] + d
                p = 3
            if p == 0:
                raise RuntimeError(
                    "Unexpected value for p=%d at position=%r." % (p, (i, j)))

            dist[i, j] = c
            pred[i, j] = p

    d = dist[len(s1), len(s2)]
    return d
'''

f2 = '''
def edit_distance_string(s1, s2):
    """
    Computes the edit distance between strings *s1* and *s2*.

    :param s1: first string
    :param s2: second string
    :return: dist, list of tuples of aligned characters
    """
    n1 = len(s1) + 1
    n2 = len(s2) + 1
    dist = numpy.full((n1, n2), n1 * n2, dtype=numpy.float64)
    pred = numpy.full(dist.shape, 0, dtype=numpy.int32)

    for i in range(0, n1):
        dist[i, 0] = i
        pred[i, 0] = 1
    for j in range(1, n2):
        dist[0, j] = j
        pred[0, j] = 2
    pred[0, 0] = -1

    for i in range(1, n1):
        for j in range(1, n2):
            c = dist[i, j]

            p = 0
            if dist[i - 1, j] + 1 < c:
                c = dist[i - 1, j] + 1
                p = 1
            if dist[i, j - 1] + 1 < c:
                c = dist[i, j - 1] + 1
                p = 2
            d = 0 if s1[i - 1] == s2[j - 1] else 1
            if dist[i - 1, j - 1] + d < c:
                c = dist[i - 1, j - 1] + d
                p = 3
            if p == 0:
                raise RuntimeError(
                    "Unexpected value for p=%d at position=%r." % (p, (i, j)))

            dist[i, j] = c
            pred[i, j] = p

    d = dist[len(s1), len(s2)]
    equals = []
    i, j = len(s1), len(s2)
    p = pred[i, j]
    while p != -1:
        if p == 3:
            equals.append((i - 1, j - 1))
            i -= 1
            j -= 1
        elif p == 2:
            j -= 1
        elif p == 1:
            i -= 1
        else:
            raise RuntimeError(
                "Unexpected value for p=%d at position=%r." % (p, (i, j)))
        p = pred[i, j]
    return d, list(reversed(equals))
'''

Visual differences: codediff ¶

%load_ext pyquickhelper

%%html
<style>
table td, table th, table tr {text-align:left !important; white-space: pre;}
</style>

This is slow due to the edit distance computation. It could be improved by a C++ implementation.

%codediff f1 f2 --verbose 1

100%|██████████████████████████████████████████████████████████████████████████████████| 47/47 [00:02<00:00, 23.05it/s]

`0`	`0`
`1`	`1`	`def edit_distance_string(s1, s2):`
`2`	`2`	`"""`
`3`	`3`	`Computes the edit distance between strings s1 and s2.`
`4`	`4`
`5`	`5`	`:param s1: first string`
`6`	`6`	`:param s2: second string`
`7`	`7`	`:return: dist, list of tuples of aligned characters`
`8`	`8`	`"""`
`9`	`9`	`n1 = len(s1) + 1`
`10`	`10`	`n2 = len(s2) + 1`
`11`	`11`	`dist = numpy.full((n1, n2), n1 * n2, dtype=numpy.float64)`
`12`	`12`	`pred = numpy.full(dist.shape, 0, dtype=numpy.int32)`
`13`	`13`
`14`	`14`	`for j in range(1, n2):` `for i in range(0, n1):`
`15`	`15`	`dist[0, j] = j` `dist[i, 0] = i`
`16`	`16`	`pred[0, j] = 2` `pred[i, 0] = 1`
`17`	`17`	`for i in range(0, n1):` `for j in range(1, n2):`
`18`	`18`	`dist[i, 0] = i` `dist[0, j] = j`
`19`	`19`	`pred[i, 0] = 1` `pred[0, j] = 2`
`20`	`20`	`pred[0, 0] = -1`
`21`	`21`
`22`	`22`	`for j in range(1, n2):` `for i in range(1, n1):`
`23`	`23`	`for i in range(1, n1):` `for j in range(1, n2):`
`24`	`24`	`c = dist[i, j]`
`25`	`25`
`26`	`26`	`p = 0`
`27`	`27`	`if dist[i - 1, j] + 1 < c:`
`28`	`28`	`c = dist[i - 1, j] + 1`
`29`	`29`	`p = 1`
`30`	`30`	`if dist[i, j - 1] + 1 < c:`
`31`	`31`	`c = dist[i, j - 1] + 1`
`32`	`32`	`p = 2`
`33`	`33`	`d = 0 if s1[i - 1] == s2[j - 1] else 1`
`34`	`34`	`if dist[i - 1, j - 1] + d < c:`
`35`	`35`	`c = dist[i - 1, j - 1] + d`
`36`	`36`	`p = 3`
`37`	`37`	`if p == 0:`
`38`	`38`	`raise RuntimeError(`
`39`	`39`	`"Unexpected value for p=%d at position=%r." % (p, (i, j)))`
`40`	`40`
`41`	`41`	`dist[i, j] = c`
`42`	`42`	`pred[i, j] = p`
`43`	`43`
`44`	`44`	`d = dist[len(s1), len(s2)]`
`45`		`return d`
	`45`	`equals = []`
	`46`	`i, j = len(s1), len(s2)`
	`47`	`p = pred[i, j]`
	`48`	`while p != -1:`
	`49`	`if p == 3:`
	`50`	`equals.append((i - 1, j - 1))`
	`51`	`i -= 1`
	`52`	`j -= 1`
	`53`	`elif p == 2:`
	`54`	`j -= 1`
	`55`	`elif p == 1:`
	`56`	`i -= 1`
	`57`	`else:`
	`58`	`raise RuntimeError(`
	`59`	`"Unexpected value for p=%d at position=%r." % (p, (i, j)))`
	`60`	`p = pred[i, j]`
	`61`	`return d, list(reversed(equals))`
`46`	`62`

%codediff f1 f2 --verbose 1 --two 1

100%|██████████████████████████████████████████████████████████████████████████████████| 47/47 [00:02<00:00, 22.99it/s]

`0`	`0`
`1`	`1`	`def edit_distance_string(s1, s2):`	`def edit_distance_string(s1, s2):`
`2`	`2`	`"""`	`"""`
`3`	`3`	`Computes the edit distance between strings s1 and s2.`	`Computes the edit distance between strings s1 and s2.`
`4`	`4`
`5`	`5`	`:param s1: first string`	`:param s1: first string`
`6`	`6`	`:param s2: second string`	`:param s2: second string`
`7`	`7`	`:return: dist, list of tuples of aligned characters`	`:return: dist, list of tuples of aligned characters`
`8`	`8`	`"""`	`"""`
`9`	`9`	`n1 = len(s1) + 1`	`n1 = len(s1) + 1`
`10`	`10`	`n2 = len(s2) + 1`	`n2 = len(s2) + 1`
`11`	`11`	`dist = numpy.full((n1, n2), n1 * n2, dtype=numpy.float64)`	`dist = numpy.full((n1, n2), n1 * n2, dtype=numpy.float64)`
`12`	`12`	`pred = numpy.full(dist.shape, 0, dtype=numpy.int32)`	`pred = numpy.full(dist.shape, 0, dtype=numpy.int32)`
`13`	`13`
`14`	`14`	`for j in range(1, n2):`	`for i in range(0, n1):`
`15`	`15`	`dist[0, j] = j`	`dist[i, 0] = i`
`16`	`16`	`pred[0, j] = 2`	`pred[i, 0] = 1`
`17`	`17`	`for i in range(0, n1):`	`for j in range(1, n2):`
`18`	`18`	`dist[i, 0] = i`	`dist[0, j] = j`
`19`	`19`	`pred[i, 0] = 1`	`pred[0, j] = 2`
`20`	`20`	`pred[0, 0] = -1`	`pred[0, 0] = -1`
`21`	`21`
`22`	`22`	`for j in range(1, n2):`	`for i in range(1, n1):`
`23`	`23`	`for i in range(1, n1):`	`for j in range(1, n2):`
`24`	`24`	`c = dist[i, j]`	`c = dist[i, j]`
`25`	`25`
`26`	`26`	`p = 0`	`p = 0`
`27`	`27`	`if dist[i - 1, j] + 1 < c:`	`if dist[i - 1, j] + 1 < c:`
`28`	`28`	`c = dist[i - 1, j] + 1`	`c = dist[i - 1, j] + 1`
`29`	`29`	`p = 1`	`p = 1`
`30`	`30`	`if dist[i, j - 1] + 1 < c:`	`if dist[i, j - 1] + 1 < c:`
`31`	`31`	`c = dist[i, j - 1] + 1`	`c = dist[i, j - 1] + 1`
`32`	`32`	`p = 2`	`p = 2`
`33`	`33`	`d = 0 if s1[i - 1] == s2[j - 1] else 1`	`d = 0 if s1[i - 1] == s2[j - 1] else 1`
`34`	`34`	`if dist[i - 1, j - 1] + d < c:`	`if dist[i - 1, j - 1] + d < c:`
`35`	`35`	`c = dist[i - 1, j - 1] + d`	`c = dist[i - 1, j - 1] + d`
`36`	`36`	`p = 3`	`p = 3`
`37`	`37`	`if p == 0:`	`if p == 0:`
`38`	`38`	`raise RuntimeError(`	`raise RuntimeError(`
`39`	`39`	`"Unexpected value for p=%d at position=%r." % (p, (i, j)))`	`"Unexpected value for p=%d at position=%r." % (p, (i, j)))`
`40`	`40`
`41`	`41`	`dist[i, j] = c`	`dist[i, j] = c`
`42`	`42`	`pred[i, j] = p`	`pred[i, j] = p`
`43`	`43`
`44`	`44`	`d = dist[len(s1), len(s2)]`	`d = dist[len(s1), len(s2)]`
`45`		`return d`
	`45`		`equals = []`
	`46`		`i, j = len(s1), len(s2)`
	`47`		`p = pred[i, j]`
	`48`		`while p != -1:`
	`49`		`if p == 3:`
	`50`		`equals.append((i - 1, j - 1))`
	`51`		`i -= 1`
	`52`		`j -= 1`
	`53`		`elif p == 2:`
	`54`		`j -= 1`
	`55`		`elif p == 1:`
	`56`		`i -= 1`
	`57`		`else:`
	`58`		`raise RuntimeError(`
	`59`		`"Unexpected value for p=%d at position=%r." % (p, (i, j)))`
	`60`		`p = pred[i, j]`
	`61`		`return d, list(reversed(equals))`
`46`	`62`

strdiff ¶

%strdiff f1 f2

def edit_distance_string(s1, s2):

"""

Computes the edit distance between strings *s1* and *s2*.

:param s1: first string

:param s2: second string

:return: dist, list of tuples of aligned characters

"""

n1 = len(s1) + 1

n2 = len(s2) + 1

dist = numpy.full((n1, n2), n1 * n2, dtype=numpy.float64)

pred = numpy.full(dist.shape, 0, dtype=numpy.int32)

for i in range(0, n1):

dist[i, 0] = i

pred[i, 0] = 1

for j in range(1, n2):

dist[0, j] = j

pred[0, j] = 2

for i in range(0, n1):

dist[i, 0] = i

pred[i, 0] = 1

pred[0, 0] = -1

for j in range(1, n2):

for i in range(1, n1):

for j in range(1, n2):

c = dist[i, j]

p = 0

if dist[i - 1, j] + 1 < c:

c = dist[i - 1, j] + 1

p = 1

if dist[i, j - 1] + 1 < c:

c = dist[i, j - 1] + 1

p = 2

d = 0 if s1[i - 1] == s2[j - 1] else 1

if dist[i - 1, j - 1] + d < c:

c = dist[i - 1, j - 1] + d

p = 3

if p == 0:

raise RuntimeError(

"Unexpected value for p=%d at position=%r." % (p, (i, j)))

dist[i, j] = c

pred[i, j] = p

d = dist[len(s1), len(s2)]

return d

equals = []

i, j = len(s1), len(s2)

p = pred[i, j]

while p != -1:

if p == 3:

equals.append((i - 1, j - 1))

i -= 1

j -= 1

elif p == 2:

j -= 1

elif p == 1:

i -= 1

else:

raise RuntimeError(

"Unexpected value for p=%d at position=%r." % (p, (i, j)))

p = pred[i, j]

return d, list(reversed(equals))

textdiff ¶

%textdiff f1 f2

populating...

<IPython.core.display.Javascript object>

Magic command to compare files¶

Two functions slighly different ¶

Visual differences: codediff ¶

strdiff ¶

textdiff ¶

Links

Contents

Information

Related Topics

This Page

Magic command to compare files¶

Two functions slighly different¶

Visual differences: codediff¶

strdiff¶

textdiff¶

Two functions slighly different ¶

Visual differences: codediff ¶

strdiff ¶

textdiff ¶