Fixed Some bugs of code_mobject.py and Text_mobject.py

[NavpreetDevpuri]

Fixed #1067
NOTE : SurroundingRectangle() only contains visible text.

class te(Scene):
    def construct(self):
        text1 = Text(" ab ", font="Consolas", size=2)
        rect1 = SurroundingRectangle(text1)
        text1.chars[1].set_color(GREEN)
        self.add(text1,rect1)

Output

class bug1(Scene):
    def construct(self):
        text1 = Text("  ab\ncd", font="Consolas", size=2)
        text2 = Text("ab\ngh", font="Consolas", size=2)
        rect1 = SurroundingRectangle(text1)
        rect2 = SurroundingRectangle(text2)
        self.play(Transform(remove_invisible_chars(text1), remove_invisible_chars(text2)), Transform(rect1, rect2))
        self.wait()

Output

Added new parameters background_stroke_width and background_stroke_color

class bug2andbug3(Scene):
    def construct(self):
        helloworldc = Code(
            "helloworldc.c",
            tab_width=4,
            background_stroke_width=1,
            background_stroke_color=WHITE,
            insert_line_no=True,
            style=Code.styles_list[4],
            background="window",
            language="cpp",
        )
        helloworldcpp = Code(
            "helloworldcpp.cpp",
            tab_width=4,
            background_stroke_width=1,
            background_stroke_color=WHITE,
            insert_line_no=True,
            style=Code.styles_list[15],
            background="window",
            language="cpp",
        )
        helloworldc.move_to(np.array([-3.6, 0, 0]))
        helloworldcpp.move_to(np.array([3.1, 0, 0]))
        self.add(helloworldc,helloworldcpp)

Output

Updated Paragraph() and added new methods set_line_to_initial_position() and set_all_lines_to_initial_positions()

class para(Scene):
    def construct(self):
        t = Paragraph('this is a awesome', 'paragraph', 'With \nNewlines', '\tWith Tabs', '  With Spaces',
                      'With Alignments',
                      'center', "left", "right")
        t.set_line_alignment("center", 7)
        t.set_line_alignment("right", 9)
        t.shift(2 * RIGHT)
        rect = SurroundingRectangle(t)
        self.add(t, rect)
        self.wait()
        self.play(t.set_all_lines_alignments, "right")
        self.play(t.set_line_to_initial_position, 4)
        self.play(t.set_all_lines_to_initial_positions)
        self.play(t.set_line_alignment, "center", 7)
        self.play(t.set_line_alignment, "right", 9)
        t.chars[0][5:7].set_color(GREEN)
        t[1][0:4].set_color(YELLOW)
        self.wait()

Output

[TonyCrane]

Good job!
But can you fix the second bug mentioned in #1067, by making Code() distinguish between space and tabs for identation.

And I think it's better to put front space characters in the first display character's position, which makes them in the first line. This can make Transform more natural.

[NavpreetDevpuri]

But can you fix the second bug mentioned in #1067, by making Code() distinguish between space and tabs for identation.
And I think it's better to put front space characters in the first display character's position, which makes them in the first line. This can make Transform more natural.

i don't get you.

i just make cairo to handle spaces or tabs (replaced by tab_width number of spaces before cairo)
and put empty Dot() as spaces at the center of previous character in text so characters indexes works correctly and transform() works correctly because Dot() are placed under some other character.

[TonyCrane]

Sorry for your failure to understand me.
The indexes work well. But there is still a little problem with Transform()
now the structure of Text() is like this:

all of the front spaces(Dot()s) are in the place of the last character(e), this cause the last character Transform to the first character(not in the same line) when TransformText.
So I think it will be better if you can put the front spaces(Dot()s which are not displayed) in the place of the first character('a' which in the same line as front spaces), just like:

I try it manually:
before:

text = Text("  ab\ncd\nef", font="Consolas", size=2)
text2 = Text("ab\n  cd\nef", font="Consolas", size=2)
self.add(text)
self.wait()
self.play(Transform(text, text2))
self.wait()

(See that? 'ab' is growing from 'f' when transforming)
after:

text = Text("  ab\ncd\nef", font="Consolas", size=2)
text2 = Text("ab\n  cd\nef", font="Consolas", size=2)

text[:2].move_to(text[2])

self.add(text)
self.wait()
self.play(Transform(text, text2))
self.wait()

[NavpreetDevpuri]

Now i get it
actually i did that i copied line and forgot to remove -1 or may be i undo for something or mistakenly did it more then one.

    def apply_space_chars(self):
        for char_index in range(self.text.__len__()):
            if self.text[char_index] == " " or self.text[char_index] == "\t" or self.text[char_index] == "\n":
                space = Dot(redius=0, fill_opacity=0, stroke_opacity=0)
                if char_index == 0:
                    #Here is the mistake i forgot to remove -1 
                    space.move_to(self.submobjects[char_index - 1].get_center())
                else:
                    space.move_to(self.submobjects[char_index - 1].get_center())
                self.submobjects.insert(char_index, space)

[TonyCrane]

focus on here.
You put all non-display characters in the position of the previous character.
But if char_index==0, the previous character is the last display character in the entire Text.

It's better to put them in the position of the first display character instead of the last one.

[NavpreetDevpuri]

Yes corrected !

[TonyCrane]

Good Job!
What about Bug 2. You can check the input code file first for identation, and replace tab_width spaces with a 'tab' before all.

[NavpreetDevpuri]

if i remove this lines

        SVGMobject.__init__(self, file_name, **config)
        #self.text = text
        #self.apply_space_chars()

It shows more smooth Transformation

But this will result into failure of indexes i.e spaces are not counted

[NavpreetDevpuri]

its better if we just accept that space characters are not counted and make Text() that way.
what do you think?

[TonyCrane]

Yes, Tranform's problems are all above the spaces.
If you can only receive spaces, but do not allow spaces to participate in Transform, it is better.
However, it is still required to calculate spaces on the index, which is a feature of Text.
For example: text=Text("ab cd")
if we don't display spaces(don't let spaces participate in Transform), but still let text[4]='c', This is the best. But it seems difficult to achieve.

[NavpreetDevpuri]

well i have an idea but for now i don't know to achieve that
the idea is to introduce a None submobject that don't participate in any of submobject's things like transformation.

[NavpreetDevpuri]

can you explain me
let text=Text("ab cd")
now my question is why text is acting as an array like text[4]='c' and why text is equivalate to text.submobject

[TonyCrane]

oops, I typed two spaces between b and c, but maybe GitHub shows only one space.

Text is a subclass of SVGMobject, so the submobjects of Text is each path(character) in the svg which cairo made.
I said text[4]='c' is just for convenience. My meaning is that text[4] points to the object with the character 'c' in text.

[TonyCrane]

@NavpreetDevpuri I solved Bug2.

def replace_identation(self):
    with open(self.file_path, 'r') as fpr:
        content = fpr.read()
    content = re.sub(' ' * self.tab_width, '\t', content)
    with open(self.file_path, 'w') as fpw:
        fpw.write(content)

add this method to Code(), and use it after ensure_valid_fild()

self.ensure_valid_file()
self.replace_identation()
self.style = self.style.lower()

You can commit it to your pr. and it will be perfect.

[NavpreetDevpuri]

@NavpreetDevpuri I solved Bug2.
content = re.sub(' ' * self.tab_width, '\t', content)

this will replace all spaces from file for example printf(" ab") will convert to printf("\tab"
you may confused about tab_width and indentation_char
tab_width defines number of spaces in indentation
there is another parameter indentation_char which defines how many spaces is equal to one indentation for example some may have only two spaces defines one indentation as for your code its four spaces equal to one indentation

we have to use indentation_char because we don't know how many spaces user defines equal to indentation

[NavpreetDevpuri]

m working on it

class te1(Scene):
    def construct(self):
        text = Text("  ab\ncd\nef", font="Consolas", size=2)
        text2 = Text("ab\n  cd\nef", font="Consolas", size=2)
        self.add(text)
        text.chars[2].set_color(YELLOW)
        self.wait()
        text.chars[0:4].set_color(RED)
        text2.chars[0:2].set_color(GREEN)
        text.set_color_by_t2c(t2c={"cd": GREEN})
        self.wait()
        # Wrong Transform(text.chars[0:4], text2.chars[0:2])
        self.play(Transform(text.get_sub_text(0, 4), text2.get_sub_text(0, 2)))
        self.wait()

Transformations are smoother and we can access chars by text.chars[] and get sub text by text.get_sub_text()

now m working on Paragraph() and Code()

[TonyCrane]

we have to use indentation_char because we don't know how many spaces user defines equal to indentation

Oh, Sorry. I didn't notice the keyword indentation_char before. This is not a bug, so sorry.

[NavpreetDevpuri]

No problem

[TonyCrane]

m working on it

Transformations are smoother and we can access chars by text.chars[] and get sub text by text.get_sub_text()

now m working on Paragraph() and Code()

But don't you think this will make Transform more complicated.
I think it is good enough now.

[NavpreetDevpuri]

But don't you think this will make Transform more complicated.

yes, but then we just have choice that

• use simple text[] to access characters without counting spaces(invisible chars)
• use text.chars[] and text.get_sub_text() counting spaces.

[NavpreetDevpuri]

# WRONG
        #self.play(Transform(text.chars[0:4], text2.chars[0:2]))
        self.play(Transform(remove_spaces_from_chars(text.chars[0:4]), remove_spaces_from_chars(text2.chars[0:2])))

Now look better remove_spaces_from_chars() for removing spaces from text.chars[]

[TonyCrane]

I still think it is too complicated.

[NavpreetDevpuri]

reopened at ManimCommunity/manim#198