60 — Search — finding items in JSON

A catalog is a list of records — dicts with a name and tags. A search loop asks for a keyword, walks the list, and keeps every record whose name matches. This guide loads the catalog from a local file or an HTTP server and matches text case-insensitively with .lower().

#Steps

1. One record has a name and a tags list. The whole catalog is a list of those dicts, saved as catalog.json:

   [
     {"name": "Red Apple", "tags": ["fruit", "sweet"]},
     {"name": "Red Rose", "tags": ["flower", "garden"]},
     {"name": "Green Tea", "tags": ["drink", "warm"]},
     {"name": "Blueberry", "tags": ["fruit", "blue"]}
   ]
   

   copyrun it →

2. Load the catalog from the file next to the program, or from a local server when the file is missing. os.path.exists picks the path; the server branch is the urlopen line from guide 55. Both json_load and loads return the same shape — a list of dicts:

   import os
   use file latest
   from json import loads
   from urllib.request import urlopen
   
   if os.path.exists("catalog.json"):
       items = json_load("catalog.json")
   else:
       url = "http://localhost:8000/catalog.json"
       items = loads(urlopen(url).read().decode("utf-8"))
   

   copyrun it →

3. "Red" in "Red Apple" is case-sensitive, so a lowercase search would miss it. .lower() copies a string in lowercase; on both sides it makes the match case-insensitive. Tags are a list, so word.lower() in item["tags"] checks the whole tag list — the same in operator on a string and a list. It prints Red Apple twice:

   name = "Red Apple"
   word = "red"
   if word.lower() in name.lower():
       show name
   
   item = {"name": "Red Apple", "tags": ["fruit", "sweet"]}
   if "sweet" in item["tags"]:
       show item["name"]
   

   copyrun it →

4. A found counter turns "no matches" into a real answer instead of silence:

   found = 0
   for item in items:
       if "red" in item["name"].lower():
           show f"{item['name']}: {', '.join(item['tags'])}"
           found = found + 1
   if found == 0:
       show "no matches"
   

   copyrun it →

5. The whole program. Save search.nme next to catalog.json:

   # search.nme — find items in a JSON catalog.
   # Run: nme r search
   # Type search, list, or quit.
   
   import os
   use file latest
   from json import loads
   from urllib.request import urlopen
   
   # Load the catalog from the local file, or from a local server.
   if os.path.exists("catalog.json"):
       items = json_load("catalog.json")
   else:
       url = "http://localhost:8000/catalog.json"
       items = loads(urlopen(url).read().decode("utf-8"))
   
   show f"catalog: {len(items)} items"
   while True:
       show "Commands: search, list, quit"
       ask command, "? "
       if command == "search":
           ask word, "Keyword? "
           found = 0
           for item in items:
               # Lowercase both sides, so Red finds red.
               name = item["name"].lower()
               if word.lower() in name or word.lower() in item["tags"]:
                   show f"{item['name']}: {', '.join(item['tags'])}"
                   found = found + 1
           if found == 0:
               show "no matches"
       elif command == "list":
           for item in items:
               show item["name"]
       elif command == "quit":
           show "Bye!"
           break
       else:
           show "Unknown command"
   

   copyrun it →

   The search checks the name and the tags, so red finds Red Apple and Red Rose by name, and blue finds Blueberry through its tag.

6. Run it and feed the commands through a pipe. search looks up red, list prints every name, and quit leaves the loop:

   printf 'search\nred\nlist\nquit\n' | nme r search
   

   copy

   catalog: 4 items
   Commands: search, list, quit
   ? Keyword? Red Apple: fruit, sweet
   Red Rose: flower, garden
   Commands: search, list, quit
   ? Red Apple
   Red Rose
   Green Tea
   Blueberry
   Commands: search, list, quit
   ? Bye!
   

   copy

   A keyword with no matches prints no matches. To use the server branch, start python3 -m http.server 8000 in the folder, rename catalog.json away, and the same program fetches the identical list over HTTP.

#Try it yourself

Search the tags only — drop the name check and match word.lower() against each item's tag list. Or print f"{found} matches" after the loop.

#What you learned

• A catalog is a list of dicts, each with a name and tags.
• os.path.exists chooses between json_load and loads(urlopen(...)).
• .lower() on both sides makes an in match case-insensitive.
• in works on both strings (name) and lists (tags).
• A found counter distinguishes no matches from an empty loop.