1
0
Fork 0
forked from fun/fun

Holiday is over, now I have fun again. No code changes. (0.38.11)

This commit is contained in:
Johannes Findeisen 2026-02-17 14:34:57 +01:00
commit c44eba681c
2 changed files with 400 additions and 0 deletions

197
docs/arrays.md Normal file
View file

@ -0,0 +1,197 @@
# Arrays in Fun
This guide focuses on arrays: creation, indexing, mutation, iteration, slicing, and common gotchas. It complements the quick overview in types.md with a deeper, exampledriven treatment.
## What is an array?
- Ordered, zeroindexed, mutable sequence of values.
- Can hold mixed types (numbers, strings, maps, arrays, …) in the same array.
- Boundschecked indexing; outofrange access is a runtime error.
## Creating arrays
```
// literals
a = [1, 2, 3]
b = ["alpha", "beta"]
empty = []
// nested
grid = [[1,2], [3,4]]
print(typeof(a)) // "array"
print(len(a)) // 3
```
Tip: Prefer squarebracket literals for clarity and performance versus building via repeated push in a hot loop.
## Indexing (0based) and assignment
```
a = [10, 20, 30]
print(a[0]) // 10
print(a[2]) // 30
// update in place
a[1] = 42
print(a) // [10, 42, 30]
```
Notes:
- Valid indices are 0..len(a)-1. Using an invalid index raises a runtime error.
- Assignment updates the existing array; references pointing to it observe the change.
## Appending, popping, inserting, removing
```
a = [1]
// append to end; returns new length
push(a, 7) // => 2, a is now [1, 7]
// pop last element; returns the removed value
v = apop(a) // v = 7, a is now [1]
// insert at index (shifts elements to the right)
a = [1, 2, 3]
insert(a, 1, 99) // a => [1, 99, 2, 3]
// remove at index (shifts left)
remove(a, 2) // a => [1, 99, 3]
```
## Slicing and concatenation
```
a = [0,1,2,3,4]
// slice(startInclusive, endExclusive)
head = slice(a, 0, 3) // [0,1,2]
mid = slice(a, 1, 4) // [1,2,3]
// concat: join two arrays
b = ["x", "y"]
ab = concat(a, b) // [0,1,2,3,4,"x","y"]
```
Slicing returns a new array. The original is unchanged.
## Iteration patterns
```
a = ["a", "b", "c"]
// indexbased loop
for i = 0; i < len(a); i = i + 1 {
print(a[i])
}
// enumerate helper (if available in your stdlib setup)
#include <utils/iter.fun> as it
for pair in it.enumerate(a) {
idx = pair[0]
val = pair[1]
print(to_string(idx) + ":" + val)
}
```
## Copying vs. referencing
Arrays are reference types. Assigning just copies the reference, not the contents:
```
orig = [1, 2]
alias = orig // points to the same array
alias[0] = 9
print(orig) // [9, 2]
// create a shallow copy via slice
copy = slice(orig, 0, len(orig))
copy[1] = 7
print(orig) // [9, 2]
print(copy) // [9, 7]
```
Shallow copies duplicate the toplevel array but not nested structures.
## Equality
```
print([1,2] == [1,2]) // true
print([1,2] == [2,1]) // false
```
Array equality compares length and elementwise equality recursively.
## Common utilities
Depending on your build/stdlib configuration, these helpers may be available:
- len(a): number of elements
- push(a, v), apop(a)
- insert(a, idx, v), remove(a, idx)
- slice(a, start, end)
- concat(a, b)
- find(a, v): index or -1
- contains(a, v): 1 or 0
Check your lib directory (e.g., lib/utils) for additional helpers.
## Error handling and bounds
```
a = [0]
// a[1] is out of range → runtime error
```
Tips:
- Guard indices: if i < 0 or i >= len(a) { /* handle */ }
- Use remove/insert carefully inside loops; indices of following items change.
## Interop with maps and strings
```
// arrays of maps
users = [ {"name":"Ada"}, {"name":"Lin"} ]
print(users[1]["name"]) // Lin
// split/join patterns depend on your stdlib
#include <utils/strings.fun> as su // adjust if present in your tree
parts = su.split("a,b,c", ",") // ["a","b","c"]
csv = su.join(parts, ",") // "a,b,c"
```
## Performance tips
- Preallocate by building from literals or chunked appends rather than onebyone in very tight loops.
- Prefer index loops over repeated remove/insert in the middle of large arrays.
- Use slice to copy only when necessary; keep references for readonly sharing.
## Examples
```
// filter even numbers
src = [0,1,2,3,4,5]
dst = []
for i = 0; i < len(src); i = i + 1 {
v = src[i]
if v % 2 == 0 { push(dst, v) }
}
print(dst) // [0,2,4]
// flatten one level
nested = [[1,2], [3], [], [4,5]]
flat = []
for i = 0; i < len(nested); i = i + 1 {
row = nested[i]
for j = 0; j < len(row); j = j + 1 {
push(flat, row[j])
}
}
print(flat) // [1,2,3,4,5]
```
## See also
- types.md — broader overview of core types with quick array examples.
- examples/ — many scripts operate on arrays; try play.fun to explore.

203
docs/maps.md Normal file
View file

@ -0,0 +1,203 @@
# Maps in Fun
This guide focuses on maps: creation, reading/writing by key, checking key presence, iterating keys/values, nesting, and common gotchas. It complements the quick overview in types.md with a deeper, exampledriven treatment.
## What is a map?
- Associative container of key → value pairs.
- Keys are typically strings; values can be any type (numbers, strings, arrays, maps, …).
- Mutable: assigning with m["k"] = v updates the map in place.
- Accessing a missing key yields nil (not an error). Use guards or has(m, key).
- Maps are not ordered; rely on keys(m) if you need a concrete list of keys.
## Creating maps
```
// literals
user = { "name": "Ada", "age": 37 }
empty = {}
print(typeof(user)) // "map"
```
Nested structures are natural and common:
```
book = {
"title": "Fun Handbook",
"meta": { "pages": 120, "isbn": "123-456" },
"tags": ["lang", "vm"]
}
print(book["meta"]["pages"]) // 120
```
## Getting and setting by key
```
profile = { "name": "Lin" }
// read existing key
print(profile["name"]) // Lin
// read missing key → nil
print(profile["email"]) // nil
// write / overwrite
profile["email"] = "lin@example.org"
profile["name"] = "Linus"
print(profile) // {"name":"Linus","email":"lin@example.org"}
```
Notes:
- Using a nonstring key is allowed only if your build/runtime supports it; most code uses string keys for portability.
- Missing keys produce nil. Compare against nil before converting or indexing:
```
v = profile["phone"]
if v == nil { print("no phone on file") }
```
## Checking key existence
Use has(m, key) to check if a key is present (returns 1 or 0):
```
cfg = { "debug": 1 }
print(has(cfg, "debug")) // 1
print(has(cfg, "port")) // 0
if has(cfg, "port") { print(cfg["port"]) } else { print("using default port") }
```
## Iterating maps
Maps are not inherently ordered. To iterate, first obtain an array of keys or values.
```
user = { "name": "Ada", "age": 38 }
// iterate known keys (explicit order you choose)
order = ["name", "age"]
for i = 0; i < len(order); i = i + 1 {
k = order[i]
print(k + " = " + to_string(user[k]))
}
// discover keys from the map (order may depend on implementation)
ks = keys(user) // -> ["name", "age"] (example)
for i = 0; i < len(ks); i = i + 1 {
k = ks[i]
print(k + ": " + to_string(user[k]))
}
// values only
vs = values(user) // -> ["Ada", 38]
for i = 0; i < len(vs); i = i + 1 { print(to_string(vs[i])) }
```
Tip:
- If you need deterministic output, either define the order array explicitly or sort the result of keys(user) using your available utilities before looping.
## Copying vs. referencing
Maps are reference types. Assigning copies the reference, not the contents:
```
orig = { "a": 1 }
alias = orig
alias["a"] = 9
print(orig["a"]) // 9
// To make a shallow copy, rebuild from keys/values
src = { "x": 1, "y": 2 }
dst = {}
ks = keys(src)
for i = 0; i < len(ks); i = i + 1 { k = ks[i]; dst[k] = src[k] }
dst["x"] = 7
print(src["x"]) // 1
print(dst["x"]) // 7
```
Shallow copies duplicate only the toplevel mapping; nested arrays/maps inside are still shared unless you clone them manually.
## Equality
```
print({"a":1,"b":2} == {"b":2,"a":1}) // true
print({"a":1} == {"a":2}) // false
```
Map equality compares sets of keys and their corresponding values for equality (order does not matter).
## Common utilities
Depending on your build/stdlib configuration, these helpers are commonly available:
- has(m, key): 1 if present, 0 otherwise
- keys(m): array of keys
- values(m): array of values (parallel to keys(m) order)
- len(a) works on arrays; for maps, prefer keys(m) then len(keys(m)) if you need a count
Check your lib or VM docs (e.g., src/vm/maps) and docs/types.md for availability and details.
## Interop with arrays and strings
```
// maps inside arrays
users = [ {"name":"Ada"}, {"name":"Lin"} ]
for i = 0; i < len(users); i = i + 1 {
print(users[i]["name"]) // Ada, Lin
}
// arrays inside maps
m = { "nums": [1,2,3] }
arr = m["nums"]
push(arr, 4)
print(m["nums"]) // [1,2,3,4]
// JSON interop is typically via lib/io/json.fun (if enabled in your build)
#include <io/json.fun> as json // adjust to your tree and build flags
s = json.stringify({"ok":1}) // "{"ok":1}"
```
## Error handling and edge cases
- Accessing a missing key returns nil. Guard before arithmetic or nested indexing.
- Mutating a map while iterating over keys you computed earlier is safe, but remember the keys array wont update automatically; recompute if needed.
- Treat map iteration order as unspecified unless your build guarantees stability. For userfacing output, specify/derive an order explicitly.
## Performance tips
- If you will access many keys, cache ks = keys(m) once and index that array in a loop rather than calling keys(m) repeatedly.
- Prefer direct writes (m["k"]=v) over repeatedly rebuilding maps in tight loops.
- When copying, copy only required keys instead of cloning the entire map if you only need a subset.
## Examples
```
// merge defaults into config (without overwriting explicit keys)
defaults = { "host":"127.0.0.1", "port":8080, "debug":0 }
cfg = { "port": 9000 }
ks = keys(defaults)
for i = 0; i < len(ks); i = i + 1 {
k = ks[i]
if !has(cfg, k) { cfg[k] = defaults[k] }
}
print(cfg) // {"port":9000,"host":"127.0.0.1","debug":0}
// index users by id
rows = [ {"id":"u1","name":"Ada"}, {"id":"u2","name":"Lin"} ]
by_id = {}
for i = 0; i < len(rows); i = i + 1 {
r = rows[i]
by_id[r["id"]] = r
}
print(by_id["u2"]["name"]) // Lin
```
## See also
- types.md — broader overview of core types with quick map examples.
- examples/ — there are example scripts involving maps; try play.fun to explore.