Table of Content

1. Introduction
2. Scanning the text
3. Looking for specific characters
4. Identifying elements
5. Processing elements

Introduction

Our goal in this post is to be able to transform this piece of json into a Python dictionary.

{"x":"y"}

Our Python dictionary should have key x and value y.

We aim to cover this without any previous compiler theory experience.

This was inspired by the Python Mauritius Usergroup’s January challenge.

Scanning the text

Our first task is to be able to go over the characters and print them.

text = '''
{"x": "y", "z":"a"}
'''


def scan_text(text):
    go = True
    current = 0

    while go:
        char = text[current]
        print(current, char.replace('\n', '<new line>'))

        if char == '}':

            go = False

        current += 1


print(scan_text(text))

0 <new line> 
1 { 
2 " 
3 x 
4 " 
5 : 
6 
7 "
8 y 
...

Looking for specific characters

Next, we want to be able to find the next occurance of a character, give us it’s index. We provide it from what index we want to search.

abcdefghijih

ex. find(2, 'i') should return 8

def look_for(from_i, char):
    go = True
    current = from_i+1
    while go:

        if text[current] == char:
            return current


        current += 1

text = 'abcdefghijih'
print(look_for(2, 'i')) # 8

We can also modify it to return the text between 2 and the index it found the first i.

def look_for(from_i, char):
    chunks = []
    go = True
    current = from_i+1
    while go:

        if text[current] == char:
            return current, ''.join(chunks)
        else:
            chunks.append(text[current])


        current += 1

print(look_for(2, 'i')) # (8, 'defgh')

This will be useful when identifying strings. When we encounter a string. We find where it ends and identifies the string.

Identifying elements

We modify the scan function to return identified text.

Our goal is to convert

text = '''
{"x": "y", "z":"a"}
'''

into ['{', 'x', ':', 'y', ',', 'z', ':', 'a', '}']

We just add a list to hold the text named chunks.

def scan_text(text):
    chunks = [] # added
    go = True
    current = 0

    while go:
        char = text[current]
        print(current, char)


        if char == '{': 
            chunks.append('{') # added

        elif char == '}':
            chunks.append('}') # added
            go = False


        elif char == '"':
            # TODO

        elif char == ":":
            chunks.append(':') # added

        if char == ',':
            chunks.append(',') # added

        current += 1
    return chunks 

For the " We just identify the next " and append the chars in between to chunks.

        elif char == '"':
            look  = look_for(current,  '"') # look from the current index till the next "
            string = look[1]
            index = look[0]

            chunks.append(string) # append identified string
            current = index # Start loop from next character. current is increased further down.

Our final function looks like this

def scan_text(text):
    chunks = []
    go = True
    current = 0

    while go:
        char = text[current]
        print(current, char)


        if char == '{':
            chunks.append('{')

        elif char == '}':
            chunks.append('}')
            go = False


        elif char == '"':
            look  = look_for(current,  '"')
            string = look[1]
            index = look[0]

            chunks.append(string)
            current = index
        elif char == ":":
            chunks.append(char)

        if char == ',':
            chunks.append(',')

        current += 1
    return chunks    

Processing elements

Now that we have a list of items, we go over the list. The simple thing would be to find : and add the element before it as the key and the element after it as the value

def process(chunks):
    go = True
    current = 0

    kv = {}
    while go:
        char = chunks[current]

        if char == ':':
            kv[chunks[current-1]] = chunks[current+1]
            current += 1 
            continue
        elif char == '}':
            go = False
        current += 1 

    return kv 

Our final code

text = '''
{"x": "y", "z":"a"}
'''

def look_for(from_i, char):
    go = True
    current = from_i+1
    text_found = []
    while go:

        if text[current] == char:
            return current, ''.join(text_found)
        else:
            text_found.append(text[current])

        current += 1
        print(current)


def scan_text(text):
    chunks = []
    go = True
    current = 0

    while go:
        char = text[current]
        print(current, char)


        if char == '{':
            chunks.append('{')

        elif char == '}':
            chunks.append('}')
            go = False


        elif char == '"':
            look  = look_for(current,  '"')
            string = look[1]
            index = look[0]

            chunks.append(string)
            current = index
        elif char == ":":
            chunks.append(char)

        if char == ',':
            chunks.append(',')

        current += 1
    return chunks    

def process(chunks):
    go = True
    current = 0

    kv = {}
    while go:
        char = chunks[current]

        if char == ':':
            kv[chunks[current-1]] = chunks[current+1]
            current += 1 
            continue
        elif char == '}':
            go = False
        current += 1 

    return kv 

elements = scan_text(text)
print(elements)
kv = process(elements)
print(kv)

It goes from this

{"x": "y", "z":"a"}

to this

['{', 'x', ':', 'y', ',', 'z', ':', 'a', '}']

then to this

{'x': 'y', 'z': 'a'}

Hope you enjoyed it!