Extended core library for Clojure/Script that emphasizes:
This lib's mostly for my own use and for advanced users that feel comfortable reading this source. Not providing much beginner-oriented documentation for this, sorry.
Quick Taoensso naming conventions:
foo - Dynamic var
foo! - Fn with side effects, or that should otherwise be used cautiously
foo? - Truthy val or fn that returns truthy val
foo!? - Fn that has side effects (or requires caution) and that return
a truthy val. Note: !?, not ?!
foo$ - Fn that's notably expensive to compute (e.g. hits db)
foo_ - Derefable val (e.g. atom, volatile, delay, etc.)
foo__ - Derefable in a derefable (e.g. delay in an atom), etc.
_ - Unnamed val
_foo - Named but unused val
?foo - Optional val (emphasize that val may be nil)
foo* - A variation of foo
(e.g. foo*
macro vs foo
fn)
foo' - ''
-foo - Public implementation detail or intermediate (e.g. uncoerced) val
foo - Val "to foo" (e.g. >sender, >host), or fn to put/coerce/transform <foo - Val "from foo" (e.g. <sender, <host), or fn to take/coerce/transform ->foo - Fn to put/coerce/transform
Type affixes may be used for clarity: <prefix>-<name> - m-users, v-users, n-users, etc. (also nusers when unambiguous) <name>-<postfix> - users-map, users-vec, user-count, etc.
Regarding name heirarchy:
When there's a significant num of syms with a meaningful hierarchy,
prefer names with descending hierarchy to emphasize structure and
related groups/functionality, e.g.:
user-add
, user-remove
, user-mod
vs
add-user
, remove-user
, mod-user
, etc.
Commit message tags (in priority order): ~ - Work-in-progress (still under development) [mod] - Modify behaviour (=> breaking), [mod!], [mod!!], etc. for attention [fix] - Fix broken behaviour (=> usu. non-breaking) [new] - Add new behaviour (=> non-breaking) [nop] - Unmodified behaviour (=> non-breaking implementation or non-code changes, etc.) [x][y] - Single commit with multiple tags (in priority order), try avoid
Example commit messages: v1.0.0 (2022-01-27) ; Tagged release [new] [#122] Add new feature x (@contributor)
Extended core library for Clojure/Script that emphasizes: * Cross platform API compatibility * Flexibility * Performance * Backwards compatibility This lib's mostly for my own use and for advanced users that feel comfortable reading this source. Not providing much beginner-oriented documentation for this, sorry. Quick Taoensso naming conventions: **foo** - Dynamic var foo! - Fn with side effects, or that should otherwise be used cautiously foo? - Truthy val or fn that returns truthy val foo!? - Fn that has side effects (or requires caution) and that return a truthy val. Note: !?, not ?! foo$ - Fn that's notably expensive to compute (e.g. hits db) foo_ - Derefable val (e.g. atom, volatile, delay, etc.) foo__ - Derefable in a derefable (e.g. delay in an atom), etc. _ - Unnamed val _foo - Named but unused val ?foo - Optional val (emphasize that val may be nil) foo* - A variation of `foo` (e.g. `foo*` macro vs `foo` fn) foo' - '' -foo - Public implementation detail or intermediate (e.g. uncoerced) val >foo - Val "to foo" (e.g. >sender, >host), or fn to put/coerce/transform <foo - Val "from foo" (e.g. <sender, <host), or fn to take/coerce/transform ->foo - Fn to put/coerce/transform Type affixes may be used for clarity: <prefix>-<name> - m-users, v-users, n-users, etc. (also nusers when unambiguous) <name>-<postfix> - users-map, users-vec, user-count, etc. Regarding name heirarchy: When there's a significant num of syms with a meaningful hierarchy, prefer names with descending hierarchy to emphasize structure and related groups/functionality, e.g.: `user-add`, `user-remove`, `user-mod` vs `add-user`, `remove-user`, `mod-user`, etc. Commit message tags (in priority order): ~ - Work-in-progress (still under development) [mod] - Modify behaviour (=> breaking), [mod!], [mod!!], etc. for attention [fix] - Fix broken behaviour (=> usu. non-breaking) [new] - Add new behaviour (=> non-breaking) [nop] - Unmodified behaviour (=> non-breaking implementation or non-code changes, etc.) [x][y] - Single commit with multiple tags (in priority order), try avoid Example commit messages: v1.0.0 (2022-01-27) ; Tagged release [new] [#122] Add new feature x (@contributor)
(-assert-unstub-val s)
(-assert-unstub-val f)
(-if-cas! atom_ old-val new-val then & [?else])
Micro optimization, mostly for cljs.
Micro optimization, mostly for cljs.
(-matching-error err)
(-matching-error c err)
(-matching-error c pattern err)
(-swap-val! atom_ k f)
Used internally by memoization utils.
Used internally by memoization utils.
(after-timeout msecs & body)
Alpha, subject to change. Returns a TimeoutFuture that will execute body after timeout. Body must be non-blocking or cheap.
Alpha, subject to change. Returns a TimeoutFuture that will execute body after timeout. Body must be non-blocking or cheap.
(ajax-lite uri
{:as opts
:keys [method params headers timeout-ms resp-type with-credentials?
xhr-pool xhr-cb-fn xhr-timeout-ms]
:or {method :get
timeout-ms 10000
resp-type :auto
xhr-pool default-xhr-pool_
xhr-timeout-ms 2500}}
callback-fn)
Alpha, subject to change. Simple, lightweight Ajax via Google Closure.
Returns nil, or resulting goog.net.XhrIo
instance if one was
immediately available.
(ajax-lite "/my-post-route" {:method :post :params {:username "Rich Hickey" :type "Awesome"} :headers {"Foo" "Bar"} :resp-type :text :timeout-ms 7000 :with-credentials? false ; Enable if using CORS (requires xhr v2+)
:xhr-pool my-xhr-pool ; goog.net.XhrIoPool
instance or delay
:xhr-cb-fn (fn [xhr]) ; Called with XhrIo
from pool when available
:xhr-timeout-ms 2500 ; Max msecs to wait on pool for XhrIo
}
(fn async-callback-fn [resp-map]
(let [{:keys [success? ?status ?error ?content ?content-type]} resp-map]
;; ?status ; ∈ #{nil 200 404 ...}, non-nil iff server responded
;; ?error ; ∈ #{nil <http-error-status-code> <exception> :timeout
:abort :http-error :exception :xhr-pool-depleted}
(js/alert (str "Ajax response: " resp-map)))))
Alpha, subject to change. Simple, lightweight Ajax via Google Closure. Returns nil, or resulting `goog.net.XhrIo` instance if one was immediately available. (ajax-lite "/my-post-route" {:method :post :params {:username "Rich Hickey" :type "Awesome"} :headers {"Foo" "Bar"} :resp-type :text :timeout-ms 7000 :with-credentials? false ; Enable if using CORS (requires xhr v2+) :xhr-pool my-xhr-pool ; `goog.net.XhrIoPool` instance or delay :xhr-cb-fn (fn [xhr]) ; Called with `XhrIo` from pool when available :xhr-timeout-ms 2500 ; Max msecs to wait on pool for `XhrIo` } (fn async-callback-fn [resp-map] (let [{:keys [success? ?status ?error ?content ?content-type]} resp-map] ;; ?status ; ∈ #{nil 200 404 ...}, non-nil iff server responded ;; ?error ; ∈ #{nil <http-error-status-code> <exception> :timeout :abort :http-error :exception :xhr-pool-depleted} (js/alert (str "Ajax response: " resp-map)))))
(as-map kvs & [kf vf])
Deprecated, prefer reduce-kvs
Deprecated, prefer `reduce-kvs`
(assert-min-encore-version min-version)
Version check for dependency conflicts, etc.
Version check for dependency conflicts, etc.
(assoc-nx m kvs)
(assoc-nx m k v)
(assoc-nx m k v & kvs)
Assocs each kv iff its key doesn't already exist.
Assocs each kv iff its key doesn't already exist.
(assoc-some m kvs)
(assoc-some m k v)
(assoc-some m k v & kvs)
Assocs each kv iff its value is not nil.
Assocs each kv iff its value is not nil.
(assoc-when m kvs)
(assoc-when m k v)
(assoc-when m k v & kvs)
Assocs each kv iff its val is truthy.
Assocs each kv iff its val is truthy.
(ba->hex-str ba)
Returns byte[] for given hex string.
Returns byte[] for given hex string.
(bench* nlaps
{:keys [nlaps-warmup nthreads as-ns?] :or {nlaps-warmup 0 nthreads 1}}
f)
Repeatedly executes fn and returns time taken to complete execution.
Repeatedly executes fn and returns time taken to complete execution.
(cache f)
(cache {:keys [size ttl-ms gc-every] :as opts} f)
Returns a cached version of given referentially transparent function f
.
Like core/memoize
but:
Often faster, depending on options.
Prevents race conditions on writes.
Supports cache invalidation by prepending args with:
:mem/del
; Delete cached item for subsequent args, returns nil.:mem/fresh
; Renew cached item for subsequent args, returns new val.Supports options:
ttl-ms
; Expire cached items after <this> many msecs.
size
; Restrict cache size to <this> many items at the next garbage
; collection (GC).
gc-every
; Run garbage collection (GC) approximately once every
; <this> many calls to cached fn. If unspecified, GC rate
; will be determined automatically based on size
.
See also defn-cached
, fmemoize
, memoize-last
.
Returns a cached version of given referentially transparent function `f`. Like `core/memoize` but: - Often faster, depending on options. - Prevents race conditions on writes. - Supports cache invalidation by prepending args with: - `:mem/del` ; Delete cached item for subsequent args, returns nil. - `:mem/fresh` ; Renew cached item for subsequent args, returns new val. - Supports options: - `ttl-ms` ; Expire cached items after <this> many msecs. - `size` ; Restrict cache size to <this> many items at the next garbage ; collection (GC). - `gc-every` ; Run garbage collection (GC) approximately once every ; <this> many calls to cached fn. If unspecified, GC rate ; will be determined automatically based on `size`. See also `defn-cached`, `fmemoize`, `memoize-last`.
(call-after-timeout msecs f)
(call-after-timeout impl_ msecs f)
Alpha, subject to change.
Returns a TimeoutFuture that will execute f
after given msecs.
Does NOT do any automatic binding conveyance.
Performance depends on the provided timer implementation (impl_
).
The default implementation offers O(logn) add, O(1) cancel, O(1) tick.
See ITimeoutImpl
for extending to arbitrary timer implementations.
Alpha, subject to change. Returns a TimeoutFuture that will execute `f` after given msecs. Does NOT do any automatic binding conveyance. Performance depends on the provided timer implementation (`impl_`). The default implementation offers O(logn) add, O(1) cancel, O(1) tick. See `ITimeoutImpl` for extending to arbitrary timer implementations.
(case-eval expr & clauses)
Like case
but evals test constants for their compile-time value.
Like `case` but evals test constants for their compile-time value.
(case-insensitive-str= s1 s2)
Returns true iff given strings are equal, ignoring case.
Returns true iff given strings are equal, ignoring case.
(catching try-expr)
(catching try-expr error-sym catch-expr)
(catching try-expr error-sym catch-expr finally-expr)
(catching try-expr error-type error-sym catch-expr finally-expr)
Cross-platform try/catch/finally.
Cross-platform try/catch/finally.
(catching-rf rf)
(catching-rf error-fn rf)
Returns wrapper around given reducing function rf
so that if rf
throws, (error-fn <thrown-error> <contextual-data>) will be called.
The default error-fn
will rethrow the original error, wrapped in
extra contextual information to aid debugging.
See also catching-xform
.
Returns wrapper around given reducing function `rf` so that if `rf` throws, (error-fn <thrown-error> <contextual-data>) will be called. The default `error-fn` will rethrow the original error, wrapped in extra contextual information to aid debugging. See also `catching-xform`.
(catching-xform xform)
(catching-xform error-fn xform)
Like catching-rf
, but applies to a transducer (xform
).
Makes debugging transductions much easier by greatly improving
the information provided in any errors thrown by xform
or the
reducing fn:
(transduce (catching-xform (comp (filter even?) (map inc))) ; Modified xform <reducing-fn> <...>)
Like `catching-rf`, but applies to a transducer (`xform`). Makes debugging transductions much easier by greatly improving the information provided in any errors thrown by `xform` or the reducing fn: (transduce (catching-xform (comp (filter even?) (map inc))) ; Modified xform <reducing-fn> <...>)
(caught-error-data & body)
Handy for error-throwing unit tests.
Handy for error-throwing unit tests.
(check-all test)
(check-all test & more)
Returns all logical false/throwing expressions (ids/forms), or nil.
Returns all logical false/throwing expressions (ids/forms), or nil.
(check-some test)
(check-some test & more)
Returns first logical false/throwing expression (id/form), or nil.
Returns first logical false/throwing expression (id/form), or nil.
(compile-if test then)
(compile-if test then else)
Evaluates test
. If it returns logical true (and doesn't throw), expands
to then
, otherwise expands to else
.
Evaluates `test`. If it returns logical true (and doesn't throw), expands to `then`, otherwise expands to `else`.
(compile-ns-filter ns-pattern)
(compile-ns-filter whitelist blacklist)
Deprecated, prefer compile-str-filter
instead.
Deprecated, prefer `compile-str-filter` instead.
(compile-str-filter spec)
Compiles given spec and returns a fast (fn conform? [?in-str]).
Spec may be:
A regex pattern. Will conform on match.
A string, in which any ""s will act as wildcards (#"."). Will conform on match.
A vector or set of regex patterns or strings. Will conform on ANY match. If you need literal "*"s, use an explicit regex pattern instead.
{:allow <allow-spec> :deny <deny-spec> :cache? <bool>}. Will conform iff allow-spec matches AND deny-spec does NOT.
Input may be: namespace strings, class names, etc. Useful as string allowlist (whitelist) and/or denylist (blacklist).
Spec examples: #{}, "", "foo.bar", "foo.bar.", #{"foo" "bar."}, {:allow #{"foo" "bar."} :deny #{"foo..bar."}}
Compiles given spec and returns a fast (fn conform? [?in-str]). Spec may be: - A regex pattern. Will conform on match. - A string, in which any "*"s will act as wildcards (#".*"). Will conform on match. - A vector or set of regex patterns or strings. Will conform on ANY match. If you need literal "*"s, use an explicit regex pattern instead. - {:allow <allow-spec> :deny <deny-spec> :cache? <bool>}. Will conform iff allow-spec matches AND deny-spec does NOT. Input may be: namespace strings, class names, etc. Useful as string allowlist (whitelist) and/or denylist (blacklist). Spec examples: #{}, "*", "foo.bar", "foo.bar.*", #{"foo" "bar.*"}, {:allow #{"foo" "bar.*"} :deny #{"foo.*.bar.*"}}
(compiling-cljs?)
Return truthy iff currently generating Cljs code.
Return truthy iff currently generating Cljs code.
(cond & clauses)
Like core/cond
but supports implicit final else
clause, and special
clause keywords for advanced behaviour:
(cond :let [x "x"] ; Establish let binding/s for remaining forms :binding [x "x"] ; Establish dynamic binding/s for remaining forms :do (println (str "x value: " x)) ; Eval expr for side effects
:if-let [y "y" z nil] "y and z were both truthy"
:if-some [y "y" z nil] "y and z were both non-nil")
:let support inspired by https://github.com/Engelberg/better-cond. Simple, flexible way to eliminate deeply-nested control flow code.
Like `core/cond` but supports implicit final `else` clause, and special clause keywords for advanced behaviour: (cond :let [x "x"] ; Establish let binding/s for remaining forms :binding [*x* "x"] ; Establish dynamic binding/s for remaining forms :do (println (str "x value: " x)) ; Eval expr for side effects :if-let [y "y" z nil] "y and z were both truthy" :if-some [y "y" z nil] "y and z were both non-nil") :let support inspired by https://github.com/Engelberg/better-cond. Simple, flexible way to eliminate deeply-nested control flow code.
(cond! & clauses)
Like cond
but throws on non-match like case
and condp
.
Like `cond` but throws on non-match like `case` and `condp`.
(conj-some)
(conj-some coll)
(conj-some coll x)
(conj-some coll x & more)
Conjoins each non-nil value.
Conjoins each non-nil value.
(conj-when)
(conj-when coll)
(conj-when coll x)
(conj-when coll x & more)
Conjoins each truthy value.
Conjoins each truthy value.
(const-ba= ba1 ba2)
Constant-time ba=
.
Useful to prevent timing attacks, etc.
Constant-time `ba=`. Useful to prevent timing attacks, etc.
(const-str= s1 s2)
Constant-time string equality checker. Useful to prevent timing attacks, etc.
Constant-time string equality checker. Useful to prevent timing attacks, etc.
(counter)
(counter init)
Returns a fast atomic Counter with init
initial int value:
Experimental 3-arity version takes an action
:
:add, :set, :set-get, :get-set, :get-add, :add-get
Returns a fast atomic Counter with `init` initial int value: - (<counter> ) -> add 1, return old val - (<counter> <n>) -> add n, return old val Experimental 3-arity version takes an `action`: :add, :set, :set-get, :get-set, :get-add, :add-get
(declare-remote & syms)
Declares given ns-qualified symbols, preserving metadata. Useful for circular dependencies.
Declares given ns-qualified symbols, preserving metadata. Useful for circular dependencies.
(defalias src-sym)
(defalias alias-sym src-sym)
(defalias alias-sym src-sym alias-attrs)
Defines a local alias for the var identified by the given qualified source symbol: (defalias my-map clojure.core/map), etc.
For Clj:
Defines a local alias for the var identified by the given qualified source symbol: (defalias my-map clojure.core/map), etc. For Clj: - Source var's metadata will be preserved (docstring, arglists, etc.). - Changes to source var's value will also be applied to alias.
(defaliases & aliases)
Bulk version of defalias
.
Takes source symbols or {:keys [alias src attrs]} maps:
(defaliases
{:alias my-map, :src map, :attrs {:doc "My map
alias"}}
{:alias my-vec, :src vec, :attrs {:doc "My vec
alias"}})
Bulk version of `defalias`. Takes source symbols or {:keys [alias src attrs]} maps: (defaliases {:alias my-map, :src map, :attrs {:doc "My `map` alias"}} {:alias my-vec, :src vec, :attrs {:doc "My `vec` alias"}})
Simple one-timeout timeout implementation provided by platform timer. O(logn) add, O(1) cancel, O(1) tick. Fns must be non-blocking or cheap. Similar efficiency to core.async timers (binary heap vs DelayQueue).
Simple one-timeout timeout implementation provided by platform timer. O(logn) add, O(1) cancel, O(1) tick. Fns must be non-blocking or cheap. Similar efficiency to core.async timers (binary heap vs DelayQueue).
(defn-cached sym cache-opts & body)
Defines a cached function.
Like (def <sym> (cache <cache-opts> <body...>)), but preserves
:arglists (arity) metadata as with defn
:
(defn-cached ^:private my-fn {:ttl-ms 500} "Does something interesting, caches resultes for 500 msecs" [n] (rand-int n))
Defines a cached function. Like (def <sym> (cache <cache-opts> <body...>)), but preserves :arglists (arity) metadata as with `defn`: (defn-cached ^:private my-fn {:ttl-ms 500} "Does something interesting, caches resultes for 500 msecs" [n] (rand-int n))
(defonce sym & args)
Like core/defonce
but supports optional docstring and attrs map.
Like `core/defonce` but supports optional docstring and attrs map.
(defstub sym)
Experimental. Declares a stub var that can be initialized from any
namespace with unstub-<stub-name>
. Decouples a var's declaration
(location) and its initialization (value). Handy for defining vars in a
shared ns from elsewhere (e.g. a private or cyclic ns).
Experimental. Declares a stub var that can be initialized from any namespace with `unstub-<stub-name>`. Decouples a var's declaration (location) and its initialization (value). Handy for defining vars in a shared ns from elsewhere (e.g. a private or cyclic ns).
(deftype-print-methods & types)
Private, used by other Taoensso libs.
Private, used by other Taoensso libs.
(deprecated & body)
Elides body when taoensso.elide-deprecated
JVM property or
TAOENSSO_ELIDE_DEPRECATED
environment variable is ∈ #{"true" "TRUE"}.
Elides body when `taoensso.elide-deprecated` JVM property or `TAOENSSO_ELIDE_DEPRECATED` environment variable is ∈ #{"true" "TRUE"}.
(dis-assoc-some m kvs)
(dis-assoc-some m k v)
(dis-assoc-some m k v & kvs)
Assocs each kv if its value is not nil, otherwise dissocs it.
Assocs each kv if its value is not nil, otherwise dissocs it.
(distinct-by keyfn coll)
Deprecated, prefer xdistinct
Deprecated, prefer `xdistinct`
(distinctv coll)
(distinctv keyfn coll)
Deprecated, prefer xdistinct
Deprecated, prefer `xdistinct`
(doto-cond [sym x] & clauses)
Cross between doto
, cond->
and as->
.
Cross between `doto`, `cond->` and `as->`.
(error-data x)
Returns data map iff x
is an error of any type on platform.
Returns data map iff `x` is an error of any type on platform.
(ex-cause ex)
Copy of core/ex-cause
(added in Clojure v1.10)
Copy of `core/ex-cause` (added in Clojure v1.10)
(ex-message ex)
Copy of core/ex-message
(added in Clojure v1.10)
Copy of `core/ex-message` (added in Clojure v1.10)
(exp-backoff n-attempt)
(exp-backoff n-attempt {:keys [min max factor] :or {factor 1000}})
Returns binary exponential backoff value for n<=36.
Returns binary exponential backoff value for n<=36.
Returns true iff any files backing the given named resources have changed since last call.
Returns true iff any files backing the given named resources have changed since last call.
(finite-num? x)
Returns true iff given a number (of standard type) that is: finite (excl. NaN and infinities).
Returns true iff given a number (of standard type) that is: finite (excl. NaN and infinities).
(float? x)
Returns true iff given a number (of standard type) that is: a fixed-precision floating-point (incl. NaN and infinities).
Returns true iff given a number (of standard type) that is: a fixed-precision floating-point (incl. NaN and infinities).
(fmemoize f)
For Clj: fastest possible memoize. Non-racey, 0-3 arity only.
For Cljs: just passes through to core/memoize
.
For Clj: fastest possible memoize. Non-racey, 0-3 arity only. For Cljs: just passes through to `core/memoize`.
(format fmt & args)
Like core/format
but:
format
!).Like `core/format` but: * Returns "" when fmt is nil rather than throwing an NPE. * Formats nil as "nil" rather than "null". * Provides ClojureScript support via goog.string.format (this has fewer formatting options than Clojure's `format`!).
(future-pool n-threads)
Returns a simple semaphore-limited wrapper of Clojure's standard future
:
(fn
[f] - Blocks to acquire a future, then executes (f) on that future.
[ ] - Blocks to acquire all futures, then immediately releases them.
Useful for blocking till all outstanding work completes.
Timeout variants are also provided.
Returns a simple semaphore-limited wrapper of Clojure's standard `future`: (fn [f] - Blocks to acquire a future, then executes (f) on that future. [ ] - Blocks to acquire all futures, then immediately releases them. Useful for blocking till all outstanding work completes. Timeout variants are also provided.
(get-dynamic-assertion-data)
Prefer get-data
Prefer `get-data`
(get-file-resource-?last-modified rname)
Returns last-modified time for file backing given named resource, or nil if file doesn't exist.
Returns last-modified time for file backing given named resource, or nil if file doesn't exist.
(get-hostname)
Returns local hostname string, or nil.
Returns local hostname string, or nil.
(get-pom-version dep-sym)
Returns POM version string for given Maven dependency, or nil.
Returns POM version string for given Maven dependency, or nil.
(get-source form env)
Returns {:keys [line column file]} given a macro's &form
, &env
vals.
Returns {:keys [line column file]} given a macro's `&form`, `&env` vals.
(get-substr s start)
(get-substr s start end)
Prefer get-substr-by-idx
Prefer `get-substr-by-idx`
(get-substr-by-idx s start-idx)
(get-substr-by-idx s start-idx end-idx)
Returns ?substring from given string.
Like subs
but:
Returns nil when requested substring would be empty.
Returns ?substring from given string. Like `subs` but: - Provides consistent clj/s behaviour. - Never throws (snaps to valid indexes). - Indexes may be -ive (=> indexed from end of string). Returns nil when requested substring would be empty.
(get-substr-by-len s start-idx)
(get-substr-by-len s start-idx sub-len)
Returns ?substring from given string.
Like get-substr-by-idx
, but takes a substring-length 3rd argument.
Returns ?substring from given string. Like `get-substr-by-idx`, but takes a substring-length 3rd argument.
(get-substring s start)
(get-substring s start length)
Prefer get-substr-by-len
Prefer `get-substr-by-len`
(get-subvec v start)
(get-subvec v start end)
Like subvec
but never throws (snaps to valid start and end indexes).
Like `subvec` but never throws (snaps to valid start and end indexes).
(get-subvector v start)
(get-subvector v start length)
Like get-subvec
but:
length
instead of end
(index).start
=> index from right of vector.Like `get-subvec` but: - Takes `length` instead of `end` (index). - -ive `start` => index from right of vector.
(get-sys-bool default prop-id env-id)
If prop-id
JVM property or env-id
environment variable are set:
true
if set value is ∈ #{"true" "1" "t" "T" "TRUE"}false
if set value is ∈ #{"false" "0" "f" "F" "FALSE"}Returns default
if neither property nor environment variable is set.
If `prop-id` JVM property or `env-id` environment variable are set: - Returns `true` if set value is ∈ #{"true" "1" "t" "T" "TRUE"} - Returns `false` if set value is ∈ #{"false" "0" "f" "F" "FALSE"} - Otherwise throws Returns `default` if neither property nor environment variable is set.
(get-truss-data)
Returns current value of dynamic assertion data.
Returns current value of dynamic assertion data.
(get-win-loc)
Returns js/window
's current location as a map.
Returns `js/window`'s current location as a map.
(have x)
(have pred (:in) x)
(have pred (:in) x & more-xs)
Takes a pred and one or more vals. Tests pred against each val, trapping errors. If any pred test fails, throws a detailed assertion error. Otherwise returns input val/vals for convenient inline-use/binding.
Respects assert value so tests can be elided from production for zero runtime costs.
Provides a small, simple, flexible feature subset to alternative tools like clojure.spec, core.typed, prismatic/schema, etc.
;; Will throw a detailed error message on invariant violation: (fn my-fn [x] (str/trim (have string? x)))
You may attach arbitrary debug info to assertion violations like:
(have string? x :data {:my-arbitrary-debug-info "foo"})
Re: use of Truss assertions within other macro bodies: Due to CLJ-865, call site information (e.g. line number) of outer macro will unfortunately be lost.
See keep-callsite
util for a workaround.
See also have?
, have!
.
Takes a pred and one or more vals. Tests pred against each val, trapping errors. If any pred test fails, throws a detailed assertion error. Otherwise returns input val/vals for convenient inline-use/binding. Respects *assert* value so tests can be elided from production for zero runtime costs. Provides a small, simple, flexible feature subset to alternative tools like clojure.spec, core.typed, prismatic/schema, etc. ;; Will throw a detailed error message on invariant violation: (fn my-fn [x] (str/trim (have string? x))) You may attach arbitrary debug info to assertion violations like: `(have string? x :data {:my-arbitrary-debug-info "foo"})` Re: use of Truss assertions within other macro bodies: Due to CLJ-865, call site information (e.g. line number) of outer macro will unfortunately be lost. See `keep-callsite` util for a workaround. See also `have?`, `have!`.
(have! x)
(have! pred (:in) x)
(have! pred (:in) x & more-xs)
Like have
but ignores assert value (so can never be elided). Useful
for important conditions in production (e.g. security checks).
Like `have` but ignores *assert* value (so can never be elided). Useful for important conditions in production (e.g. security checks).
(have!? x)
(have!? pred (:in) x)
(have!? pred (:in) x & more-xs)
Specialized cross between have?
and have!
. Not used often but can be
handy for semantic clarification and/or to improve multi-val performance
when the return vals aren't necessary.
WARNING: Do NOT use in :pre/:post conds since those are ALWAYS subject
to assert, directly contradicting the intention of the bang (!
) here.
Specialized cross between `have?` and `have!`. Not used often but can be handy for semantic clarification and/or to improve multi-val performance when the return vals aren't necessary. **WARNING**: Do NOT use in :pre/:post conds since those are ALWAYS subject to *assert*, directly contradicting the intention of the bang (`!`) here.
(have? x)
(have? pred (:in) x)
(have? pred (:in) x & more-xs)
Like have
but returns true
on successful tests. In particular, this
can be handy for use with :pre/:post conditions. Compare:
(fn my-fn [x] {:post [(have nil? %)]} nil) ; {:post [nil]} FAILS
(fn my-fn [x] {:post [(have? nil? %)]} nil) ; {:post [true]} passes as intended
Like `have` but returns `true` on successful tests. In particular, this can be handy for use with :pre/:post conditions. Compare: (fn my-fn [x] {:post [(have nil? %)]} nil) ; {:post [nil]} FAILS (fn my-fn [x] {:post [(have? nil? %)]} nil) ; {:post [true]} passes as intended
(hex-str->ba s)
Returns hex string for given byte[].
Returns hex string for given byte[].
(ident-hex-str obj)
Returns hex string of given Object's identityHashCode
(e.g. "0x5eeb49f2").
Returns hex string of given Object's `identityHashCode` (e.g. "0x5eeb49f2").
(if-let bindings then)
(if-let bindings then else)
Like core/if-let
but can bind multiple values for then
iff all tests
are truthy, supports internal unconditional :let
s.
Like `core/if-let` but can bind multiple values for `then` iff all tests are truthy, supports internal unconditional `:let`s.
(if-not test-or-bindings then)
(if-not test-or-bindings then else)
Like core/if-not
but acts like if-let
when given a binding vector
as test expr.
Like `core/if-not` but acts like `if-let` when given a binding vector as test expr.
(if-some bindings then)
(if-some bindings then else)
Like core/if-some
but can bind multiple values for then
iff all tests
are non-nil, supports internal unconditional :let
s.
Like `core/if-some` but can bind multiple values for `then` iff all tests are non-nil, supports internal unconditional `:let`s.
(instance! class arg & {:as details})
If (instance? class arg) is true, returns arg.
Otherwise throws runtime ExceptionInfo
with unexpected-arg!
.
See unexpected-arg!
for more info.
If (instance? class arg) is true, returns arg. Otherwise throws runtime `ExceptionInfo` with `unexpected-arg!`. See `unexpected-arg!` for more info.
(int? x)
Returns true iff given a number (of standard type) that is: a fixed-precision integer.
Returns true iff given a number (of standard type) that is: a fixed-precision integer.
(interleave-all)
(interleave-all c1)
(interleave-all c1 c2)
(interleave-all c1 c2 & colls)
Greedy version of interleave
.
Greedy version of `interleave`.
(interns-overview)
(interns-overview ns)
Returns {:keys [public private impl test]}, with each key mapped to an alphabetical list of the relevant vars in given namespace.
"impl" vars are public vars with names that begin with "-" or "_", a naming convention commonly used to indicate vars intended to be treated as private implementation details even when public.
Returns {:keys [public private impl test]}, with each key mapped to an alphabetical list of the relevant vars in given namespace. "impl" vars are public vars with names that begin with "-" or "_", a naming convention commonly used to indicate vars intended to be treated as private implementation details even when public.
(into-all to from)
(into-all to from & more)
Like into
but supports multiple "from"s.
Like `into` but supports multiple "from"s.
(into-str & xs)
Simple Hiccup-like string templating to complement Tempura.
Simple Hiccup-like string templating to complement Tempura.
(is! x)
(is! pred x)
(is! pred x data)
Lightweight have!
that provides less diagnostic info.
Lightweight `have!` that provides less diagnostic info.
(tf-cancel! _)
Returns true iff the timeout was successfully cancelled (i.e. was previously pending).
Returns true iff the timeout was successfully cancelled (i.e. was previously pending).
(tf-cancelled? _)
Returns true iff the timeout is cancelled.
Returns true iff the timeout is cancelled.
(tf-done? _)
Returns true iff the timeout is not pending (i.e. has a completed result or is cancelled).
Returns true iff the timeout is not pending (i.e. has a completed result or is cancelled).
(tf-pending? _)
Returns true iff the timeout is pending.
Returns true iff the timeout is pending.
(tf-poll _)
Returns :timeout/pending, :timeout/cancelled, or the timeout's completed result.
Returns :timeout/pending, :timeout/cancelled, or the timeout's completed result.
(tf-state _)
Returns a map of timeout's public state.
Returns a map of timeout's public state.
(keep-callsite & body)
CLJ-865 unfortunately means that it's currently not possible for an inner macro to access the &form metadata of an outer macro.
This means that inner macros lose call site information like the line number of the outer macro.
This util offers a workaround for macro authors:
(defmacro my-macro1 [x] (truss/have ~x)) ; W/o call site info (defmacro my-macro2 [x] (keep-callsite
(truss/have ~x))) ; With call site info
CLJ-865 unfortunately means that it's currently not possible for an inner macro to access the &form metadata of an outer macro. This means that inner macros lose call site information like the line number of the outer macro. This util offers a workaround for macro authors: (defmacro my-macro1 [x] `(truss/have ~x)) ; W/o call site info (defmacro my-macro2 [x] (keep-callsite `(truss/have ~x))) ; With call site info
(keys-by f coll)
Returns {(f x) x} map for xs in coll
.
Returns {(f x) x} map for xs in `coll`.
(limiter specs)
(limiter opts specs)
Takes {<spec-id> [<n-max-reqs> <msecs-window>]}, and returns a rate limiter (fn check-limits! [req-id]) -> nil (all limits pass), or [<worst-spec-id> <worst-backoff-msecs> {<spec-id> <backoff-msecs>}].
Limiter fn commands: :rl/peek <req-id> - Check limits w/o side effects. :rl/reset <req-id> - Reset all limits for given req-id.
Takes {<spec-id> [<n-max-reqs> <msecs-window>]}, and returns a rate limiter (fn check-limits! [req-id]) -> nil (all limits pass), or [<worst-spec-id> <worst-backoff-msecs> {<spec-id> <backoff-msecs>}]. Limiter fn commands: :rl/peek <req-id> - Check limits w/o side effects. :rl/reset <req-id> - Reset all limits for given req-id.
(limiter* specs)
(limiter* opts specs)
Experimental. Like limiter
but returns [<state_> <limiter>].
Experimental. Like `limiter` but returns [<state_> <limiter>].
(load-edn-config {:as opts
:keys [default prop env res auto-env? error-data]
:or {auto-env? true}})
Attempts to read config as EDN from the following (in descending order):
Returns nil, or {:config <loaded-config>, :source <source-of-loaded-config>}.
Throws if value/resource cannot be successfully read as EDN. The read EDN value must be a map, or a symbol that resolves to a map.
Useful for libraries, etc. that want to provide easily modified config. Used by Timbre, Carmine, etc.
Options:
env
will be provided automatically based on prop
Attempts to read config as EDN from the following (in descending order): 1. JVM property, if opt is provided and value is present. 2. Env var, if opt is provided and value is present. 3. Resource file, if opt is provided and resource is present. Returns nil, or {:config <loaded-config>, :source <source-of-loaded-config>}. Throws if value/resource cannot be successfully read as EDN. The read EDN value must be a map, or a symbol that resolves to a map. Useful for libraries, etc. that want to provide easily modified config. Used by Timbre, Carmine, etc. Options: - default ; Default config map into which a nested merge will be done - prop ; Name of JVM property to check (e.g. "taoensso.timbre.config.edn") - env ; Name of Env var to check (e.g. "TAOENSSO_TIMBRE_CONFIG_EDN") - res ; Name of resource file to check (e.g. "taoensso.timbre.config.edn") - auto-env? ; If true, `env` will be provided automatically based on `prop` - error-data ; Optional map to be added to ex-data if load fails
(map-kvs kf vf m)
Deprecated, prefer reduce-kv
Deprecated, prefer `reduce-kv`
(mapply f & args)
Like apply
but calls seq-kvs
on final arg.
Like `apply` but calls `seq-kvs` on final arg.
(memoize f)
(memoize ttl-ms f)
(memoize size ttl-ms f)
Alternative way to call cache
, provided mostly for back compatibility.
See cache
docstring for details.
Alternative way to call `cache`, provided mostly for back compatibility. See `cache` docstring for details.
(memoize-last f)
Like core/memoize
but only caches the fn's most recent call.
Great for Reactjs render op caching on mobile devices, etc.
Like `core/memoize` but only caches the fn's most recent call. Great for Reactjs render op caching on mobile devices, etc.
(merge & maps)
Like core/merge
but faster, supports :swap/dissoc
rvals.
Like `core/merge` but faster, supports `:swap/dissoc` rvals.
(merge-with f & maps)
Like core/merge-with
but faster, supports :swap/dissoc
rvals.
Like `core/merge-with` but faster, supports `:swap/dissoc` rvals.
(ms & {:as opts :keys [years months weeks days hours mins secs msecs ms]})
Returns ~number of milliseconds in period defined by given args.
Returns ~number of milliseconds in period defined by given args.
(msecs & opts)
Compile-time version of ms
Compile-time version of `ms`
(name-with-attrs sym args)
(name-with-attrs sym args attrs-merge)
Given a symbol and args, returns [<name-with-attrs-meta> <args> <attrs>]
with support for defn
style ?docstring
and ?attrs-map
.
Given a symbol and args, returns [<name-with-attrs-meta> <args> <attrs>] with support for `defn` style `?docstring` and `?attrs-map`.
(nanoid)
(nanoid len)
Experimental. Optimized variant of secure-rand-id
that returns Nano IDs
as in https://github.com/ai/nanoid.
Experimental. Optimized variant of `secure-rand-id` that returns Nano IDs as in https://github.com/ai/nanoid.
(nested-merge & maps)
Like merge
but does nested merging.
Like `merge` but does nested merging.
(nested-merge-with f & maps)
Like merge-with
but does nested merging.
Like `merge-with` but does nested merging.
(nnil)
(nnil x)
(nnil x y)
(nnil x y z)
(nnil x y z & more)
Returns first non-nil arg, or nil.
Returns first non-nil arg, or nil.
(norm-str s)
(norm-str form s)
Given a Unicode string, returns the normalized de/composed form. It's often a good idea to normalize strings before exchange or storage, especially if you're going to be querying against those string.
form
is ∈ #{:nfc :nfkc :nfd :nfkd <java.text.NormalizerForm>}.
Defaults to :nfc as per W3C recommendation.
Given a Unicode string, returns the normalized de/composed form. It's often a good idea to normalize strings before exchange or storage, especially if you're going to be querying against those string. `form` is ∈ #{:nfc :nfkc :nfd :nfkd <java.text.NormalizerForm>}. Defaults to :nfc as per W3C recommendation.
(norm-word-breaks s)
Converts all word breaks of any form and length (including line breaks of any form, tabs, spaces, etc.) to a single regular space.
Converts all word breaks of any form and length (including line breaks of any form, tabs, spaces, etc.) to a single regular space.
(now-nano)
Uses window context as epoch, Ref. http://goo.gl/mWZWnR
Uses window context as epoch, Ref. http://goo.gl/mWZWnR
(oget k)
(oget o k)
(oget o k not-found)
Like get
for JS objects.
Like `get` for JS objects.
(oget-in ks)
(oget-in o ks)
(oget-in o ks not-found)
Like get-in
for JS objects.
Like `get-in` for JS objects.
(oset-in o ks v)
Experimental. Like assoc-in
for JS objects.
Experimental. Like `assoc-in` for JS objects.
(parse-query-params s & [keywordize? encoding])
Based on ring-codec/form-decode
.
Based on `ring-codec/form-decode`.
(pnum? x)
Returns true iff given number in unsigned unit proportion interval ∈ℝ[0,1].
Returns true iff given number in unsigned unit proportion interval ∈ℝ[0,1].
(pr-edn x)
(pr-edn _opts x)
Prints arg to an edn string readable with read-edn
.
Prints arg to an edn string readable with `read-edn`.
(pre-cache n-capacity f)
(pre-cache n-capacity fp-or-n-threads f)
Given a nullary fn f
that is non-idempotent and free of side effects,
returns a wrapped version of f
that asynchronously maintains a cache
of up to n-capacity
pre-computed return values of (f).
Useful when f
is expensive & may be called in a spikey fashion,
e.g. ideal for cryptographic key generators.
Given a nullary fn `f` that is non-idempotent and free of side effects, returns a wrapped version of `f` that asynchronously maintains a cache of up to `n-capacity` pre-computed return values of (f). Useful when `f` is expensive & may be called in a spikey fashion, e.g. ideal for cryptographic key generators.
(preserve-reduced rf)
As core/preserving-reduced
.
As `core/preserving-reduced`.
(pull-val! atom_ k)
(pull-val! atom_ k not-found)
Removes and returns value mapped to key.
Removes and returns value mapped to key.
(qb spec form)
(qb spec form & more)
Simple util to benchmark/compare runtime of given form/s.
Runs sets of laps for each given form, recording the total runtime of each set. Returns the the total runtime in msecs of the fastest set of laps for each form.
(quick-bench [<num-sets> <num-laps>] <form1> <form2> <...>) => [<total runtime msecs of fastest set of laps for form1> <total runtime msecs of fastest set of laps for form2> <...>]
Total number of runs for each form is: num-sets
* num-laps
If omitted, the default num-sets
is 6 (to include warmup):
(quick-bench <num-laps> <form1> <form2> <...>)
Example (comparing runtime of first
and nth
against vector):
(let [v [:a]] (quick-bench 1e6 (first v) (nth v 0))) => [67.43 39.05]
Simple util to benchmark/compare runtime of given form/s. Runs sets of laps for each given form, recording the total runtime of each set. Returns the the total runtime in msecs of the fastest set of laps for each form. (quick-bench [<num-sets> <num-laps>] <form1> <form2> <...>) => [<total runtime msecs of fastest set of laps for form1> <total runtime msecs of fastest set of laps for form2> <...>] Total number of runs for each form is: `num-sets` * `num-laps` If omitted, the default `num-sets` is 6 (to include warmup): (quick-bench <num-laps> <form1> <form2> <...>) Example (comparing runtime of `first` and `nth` against vector): (let [v [:a]] (quick-bench 1e6 (first v) (nth v 0))) => [67.43 39.05]
(queue)
(queue coll)
Returns a PersistentQueue.
Returns a PersistentQueue.
(quick-bench spec form)
(quick-bench spec form & more)
Simple util to benchmark/compare runtime of given form/s.
Runs sets of laps for each given form, recording the total runtime of each set. Returns the the total runtime in msecs of the fastest set of laps for each form.
(quick-bench [<num-sets> <num-laps>] <form1> <form2> <...>) => [<total runtime msecs of fastest set of laps for form1> <total runtime msecs of fastest set of laps for form2> <...>]
Total number of runs for each form is: num-sets
* num-laps
If omitted, the default num-sets
is 6 (to include warmup):
(quick-bench <num-laps> <form1> <form2> <...>)
Example (comparing runtime of first
and nth
against vector):
(let [v [:a]] (quick-bench 1e6 (first v) (nth v 0))) => [67.43 39.05]
Simple util to benchmark/compare runtime of given form/s. Runs sets of laps for each given form, recording the total runtime of each set. Returns the the total runtime in msecs of the fastest set of laps for each form. (quick-bench [<num-sets> <num-laps>] <form1> <form2> <...>) => [<total runtime msecs of fastest set of laps for form1> <total runtime msecs of fastest set of laps for form2> <...>] Total number of runs for each form is: `num-sets` * `num-laps` If omitted, the default `num-sets` is 6 (to include warmup): (quick-bench <num-laps> <form1> <form2> <...>) Example (comparing runtime of `first` and `nth` against vector): (let [v [:a]] (quick-bench 1e6 (first v) (nth v 0))) => [67.43 39.05]
(rate-limiter* specs)
Deprecated, prefer limiter
Deprecated, prefer `limiter`
(rcompare x y)
Reverse comparator.
Reverse comparator.
(read-edn s)
(read-edn opts s)
Attempts to pave over differences in:
clojure.edn/read-string
, clojure.tools.edn/read-string
,
cljs.reader/read-string
, cljs.tools.reader/read-string
.
cljs.reader
in particular can be a pain.
Attempts to pave over differences in: `clojure.edn/read-string`, `clojure.tools.edn/read-string`, `cljs.reader/read-string`, `cljs.tools.reader/read-string`. `cljs.reader` in particular can be a pain.
(reduce-indexed rf init coll)
Like reduce
but takes (rf [acc idx in]) with idx as in map-indexed
.
As reduce-kv
against vector coll, but works on any seqable coll type.
Like `reduce` but takes (rf [acc idx in]) with idx as in `map-indexed`. As `reduce-kv` against vector coll, but works on any seqable coll type.
(reduce-kvs rf init kvs)
Like reduce-kv
but takes a flat sequence of kv pairs.
Like `reduce-kv` but takes a flat sequence of kv pairs.
(reduce-n rf init end)
(reduce-n rf init start end)
(reduce-n rf init start end step)
(reduce-obj f init o)
Like reduce-kv
but for JavaScript objects.
Like `reduce-kv` but for JavaScript objects.
(reduce-top n rf init coll)
(reduce-top n keyfn rf init coll)
(reduce-top n keyfn cmp rf init coll)
Reduces the top n
items from coll
of N items.
Clj impln is O(N.logn) vs O(N.logN) for (take n (sort-by ...)).
Reduces the top `n` items from `coll` of N items. Clj impln is O(N.logn) vs O(N.logN) for (take n (sort-by ...)).
(reduce-zip rf init xs ys)
Reduces given sequential xs and ys as pairs (e.g. key-val pairs). Calls (rf acc x y) for each sequential pair.
Useful, among other things, as a more flexible version of zipmap
.
Reduces given sequential xs and ys as pairs (e.g. key-val pairs). Calls (rf acc x y) for each sequential pair. Useful, among other things, as a more flexible version of `zipmap`.
(rename-keys replacements m)
Returns a map like the one given, replacing keys using the given {<old-new> <new-key>} replacements. O(min(n_replacements, n_m)).
Returns a map like the one given, replacing keys using the given {<old-new> <new-key>} replacements. O(min(n_replacements, n_m)).
(repeatedly-into coll n f)
Like repeatedly
but faster and conj
s items into given collection.
Like `repeatedly` but faster and `conj`s items into given collection.
(reset!? atom_ val)
Atomically swaps value of atom_
to val
and returns
true iff the atom's value changed. See also reset-in!?
.
Atomically swaps value of `atom_` to `val` and returns true iff the atom's value changed. See also `reset-in!?`.
(reset-in! atom_ val)
(reset-in! atom_ ks val)
(reset-in! atom_ ks not-found val)
Like reset!
but supports update-in
semantics, returns <old-key-val>.
Like `reset!` but supports `update-in` semantics, returns <old-key-val>.
(reset-in!? atom_ val)
(reset-in!? atom_ ks val)
(reset-in!? atom_ ks not-found val)
Like reset-in!
but returns true iff the atom's value changed.
Like `reset-in!` but returns true iff the atom's value changed.
(reset-val! atom_ k val)
(reset-val! atom_ k not-found val)
Like reset-in!
but optimized for single-key case.
Like `reset-in!` but optimized for single-key case.
(reset-val!? atom_ k new-val)
Like reset-in!?
but optimized for single-key case.
Like `reset-in!?` but optimized for single-key case.
(ring-redirect-resp url)
(ring-redirect-resp kind url)
(ring-redirect-resp kind url flash)
(rnum? x)
Returns true iff given number in signed unit proportion interval ∈ℝ[-1,1].
Returns true iff given number in signed unit proportion interval ∈ℝ[-1,1].
(rolling-counter msecs)
Experimental. Returns a RollingCounter that you can:
msecs
window and return RollingCounter.msecs
window.Experimental. Returns a RollingCounter that you can: - Invoke to increment count in last `msecs` window and return RollingCounter. - Deref to return count in last `msecs` window.
(rolling-list nmax)
(rolling-list nmax {:keys [init-val]})
Returns a stateful fn of 2 arities:
(fn [ ]) => Returns current array in O(n).
(fn [x]) => Adds x
to right of list, maintaining length <~ nmax
.
Returns nil. Very fast (faster than rolling-vector
).
Useful for maintaining limited-length histories, etc.
See also rolling-vector
.
Returns a stateful fn of 2 arities: (fn [ ]) => Returns current array in O(n). (fn [x]) => Adds `x` to right of list, maintaining length <~ `nmax`. Returns nil. Very fast (faster than `rolling-vector`). Useful for maintaining limited-length histories, etc. See also `rolling-vector`.
(rolling-vector nmax)
(rolling-vector nmax {:keys [gc-every init-val] :or {gc-every 16000.0}})
Returns a stateful fn of 2 arities:
(fn [ ]) => Returns current sub/vector in O(1).
(fn [x]) => Adds x
to right of sub/vector, maintaining length <= nmax
.
Returns current sub/vector.
Useful for maintaining limited-length histories, etc.
See also rolling-list
(Clj only).
Returns a stateful fn of 2 arities: (fn [ ]) => Returns current sub/vector in O(1). (fn [x]) => Adds `x` to right of sub/vector, maintaining length <= `nmax`. Returns current sub/vector. Useful for maintaining limited-length histories, etc. See also `rolling-list` (Clj only).
(satisfies! protocol arg & {:as details})
If (satisfies? protocol arg) is true, returns arg.
Otherwise throws runtime ExceptionInfo
with unexpected-arg!
.
See unexpected-arg!
for more info.
If (satisfies? protocol arg) is true, returns arg. Otherwise throws runtime `ExceptionInfo` with `unexpected-arg!`. See `unexpected-arg!` for more info.
(sb-append str-builder s)
(sb-append str-builder s & more)
For cross-platform string building
For cross-platform string building
(secure-rand-bytes size)
Returns size
random bytes using secure-rng
or js/window.crypto
.
Returns `size` random bytes using `secure-rng` or `js/window.crypto`.
(secure-rand-id alphabet len)
Experimental.
Given alphabet
(a string of possible characters), returns a securely
random string of length len
.
Trying to do this the obvious/naive way (by repeatedly generating a secure random number and mapping it to an alphabet character with modulo) actually introduces bias into ids that can be exploited by an attacker.
The method used here is designed to eliminate that bias. Based on https://bit.ly/3dtYv73.
Experimental. Given `alphabet` (a string of possible characters), returns a securely random string of length `len`. Trying to do this the obvious/naive way (by repeatedly generating a secure random number and mapping it to an alphabet character with modulo) actually introduces bias into ids that can be exploited by an attacker. The method used here is designed to eliminate that bias. Based on https://bit.ly/3dtYv73.
(secure-rng)
Returns a thread-local java.security.SecureRandom
.
Favours security over performance. Automatically re-seeds occasionally.
May block while waiting on system entropy!
Returns a thread-local `java.security.SecureRandom`. Favours security over performance. Automatically re-seeds occasionally. May block while waiting on system entropy!
(secure-rng-mock!!! long-seed)
Returns INSECURE java.security.SecureRandom
mock instance backed by
a seeded deterministic java.util.Random
. Useful for testing, etc.
Returns **INSECURE** `java.security.SecureRandom` mock instance backed by a seeded deterministic `java.util.Random`. Useful for testing, etc.
(select-nested-keys src-map key-spec)
Like select-keys
but supports nested key spec:
(select-nested-keys
{:a :A :b :B :c {:c1 :C1 :c2 :C2} :d :D} ; src-map
[:a {:c [:c1], :d [:d1 :d2]}]) ; key-spec
=> {:a :A, :c {:c1 :C1}, :d :D}
Note that as with the {:d [:d1 :d2]}
spec in the example above,
if spec expects a nested map but the actual value is not a map,
the actual value will be included in output as-is.
Has the same behaviour as select-keys
when key-spec
is a
simple vector of keys.
Like `select-keys` but supports nested key spec: (select-nested-keys {:a :A :b :B :c {:c1 :C1 :c2 :C2} :d :D} ; `src-map` [:a {:c [:c1], :d [:d1 :d2]}]) ; `key-spec` => {:a :A, :c {:c1 :C1}, :d :D} Note that as with the `{:d [:d1 :d2]}` spec in the example above, if spec expects a nested map but the actual value is not a map, the actual value will be included in output as-is. Has the same behaviour as `select-keys` when `key-spec` is a simple vector of keys.
(session-swap rreq rresp f & args)
Small util to help correctly manage (modify) funtional sessions. Please use this when writing Ring middleware! It's so easy to get this wrong and end up with subtle, tough-to-diagnose issues.
Small util to help correctly manage (modify) funtional sessions. Please use this when writing Ring middleware! It's *so* easy to get this wrong and end up with subtle, tough-to-diagnose issues.
(simple-date-format pattern & [{:keys [locale timezone] :as opts}])
Returns a thread-local java.text.SimpleDateFormat
.
Returns a thread-local `java.text.SimpleDateFormat`.
Like slurp-resource
but caches slurps against file's last-modified udt.
Like `slurp-resource` but caches slurps against file's last-modified udt.
(slurp-resource rname)
Returns slurped named resource on classpath, or nil when resource not found.
Returns slurped named resource on classpath, or nil when resource not found.
(sortv coll)
(sortv comparator coll)
(sortv ?keyfn comparator coll)
Like core/sort
but:
comparator
can be :asc
, :desc
, or an arbitrary comparator.keyfn
may be provided, as in core/sort-by
.Like `core/sort` but: - Returns a vector. - `comparator` can be `:asc`, `:desc`, or an arbitrary comparator. - An optional `keyfn` may be provided, as in `core/sort-by`.
(str-?index s substr)
(str-?index s substr start-idx)
(str-?index s substr start-idx last?)
For cross-platform string building
For cross-platform string building
(str-join coll)
(str-join separator coll)
(str-join separator xform coll)
Faster, transducer-based generalization of clojure.string/join
with xform
support.
Faster, transducer-based generalization of `clojure.string/join` with `xform` support.
(str-join-once separator coll)
Like string/join
but skips duplicate separators.
Like `string/join` but skips duplicate separators.
(str-replace s match replacement)
Like str/replace
but provides consistent clj/s behaviour.
Workaround for http://dev.clojure.org/jira/browse/CLJS-794, http://dev.clojure.org/jira/browse/CLJS-911.
Note that ClojureScript 1.7.145 introduced a partial fix for CLJS-911. A full fix could unfortunately not be introduced w/o breaking compatibility with the previously incorrect behaviour. CLJS-794 also remains unresolved.
Like `str/replace` but provides consistent clj/s behaviour. Workaround for http://dev.clojure.org/jira/browse/CLJS-794, http://dev.clojure.org/jira/browse/CLJS-911. Note that ClojureScript 1.7.145 introduced a partial fix for CLJS-911. A full fix could unfortunately not be introduced w/o breaking compatibility with the previously incorrect behaviour. CLJS-794 also remains unresolved.
(submap? m sub)
Returns true iff sub
is a (possibly nested) submap of m
,
i.e. iff every (nested) value in sub
has the same (nested) value in m
.
Warning: uses stack recursion, so supports only limited nesting.
Returns true iff `sub` is a (possibly nested) submap of `m`, i.e. iff every (nested) value in `sub` has the same (nested) value in `m`. Warning: uses stack recursion, so supports only limited nesting.
(substr s start-idx & [?max-len])
Deprecated, prefer get-substr-by-idx
or get-substr-by-len
Deprecated, prefer `get-substr-by-idx` or `get-substr-by-len`
(subvec* v start-idx & [?max-len])
Deprecated, prefer get-subvec
or get-subvector
Deprecated, prefer `get-subvec` or `get-subvector`
(swap-in! atom_ f)
(swap-in! atom_ ks f)
(swap-in! atom_ ks not-found f)
Like swap!
but supports update-in
semantics,
returns <new-key-val> or <swapped-return-val>.
Like `swap!` but supports `update-in` semantics, returns <new-key-val> or <swapped-return-val>.
(swap-in!* atom_ f)
(swap-in!* atom_ ks f)
(swap-in!* atom_ ks not-found f)
Deprecated, prefer swap-in!
with swapped
return value.
Deprecated, prefer `swap-in!` with `swapped` return value.
(swap-val! atom_ k f)
(swap-val! atom_ k not-found f)
Like swap-in!
but optimized for single-key case.
Like `swap-in!` but optimized for single-key case.
(swap-val!* atom_ k f)
(swap-val!* atom_ k not-found f)
Deprecated, prefer swap-val!
with swapped
return value.
Deprecated, prefer `swap-val!` with `swapped` return value.
(test-fixtures fixtures-map)
Given a {:before ?(fn []) :after ?(fn [])} map, returns cross-platform
test fixtures for use by both clojure.test
and cljs.test
:
(let [f (test-fixtures {:before (fn [] (test-setup))})] (clojure.test/use-fixtures :once f) (cljs.test/use-fixtures :once f))
Given a {:before ?(fn []) :after ?(fn [])} map, returns cross-platform test fixtures for use by both `clojure.test` and `cljs.test`: (let [f (test-fixtures {:before (fn [] (test-setup))})] (clojure.test/use-fixtures :once f) (cljs.test/use-fixtures :once f))
(thread-local & body)
Given a body that returns an initial value for the current thread,
returns a ThreadLocal
proxy that can be derefed to get the current
thread's current value.
Commonly used to achieve thread safety during Java interop.
In the common case, body
will be a call to some Java constructor
that returns a non-thread-safe instance.
Example:
(def thread-local-simple-date-format_
"Deref to return a thread-local SimpleDateFormat
"
(thread-local (SimpleDateFormat. "yyyy-MM-dd")))
(.format @thread-local-simple-date-format_ (Date.)) => "2023-01-24"
NB: don't pass the derefed value to other threads!
Given a body that returns an initial value for the current thread, returns a `ThreadLocal` proxy that can be derefed to get the current thread's current value. Commonly used to achieve thread safety during Java interop. In the common case, `body` will be a call to some Java constructor that returns a non-thread-safe instance. Example: (def thread-local-simple-date-format_ "Deref to return a thread-local `SimpleDateFormat`" (thread-local (SimpleDateFormat. "yyyy-MM-dd"))) (.format @thread-local-simple-date-format_ (Date.)) => "2023-01-24" NB: don't pass the derefed value to other threads!
(thread-local* init-val-fn)
Low-level, see thread-local
instead.
Low-level, see `thread-local` instead.
(thread-local-proxy & body)
Low-level, see thread-local
instead.
Low-level, see `thread-local` instead.
(throws form)
(throws c form)
(throws c pattern form)
Like throws?
, but returns ?matching-error instead of true/false.
Like `throws?`, but returns ?matching-error instead of true/false.
(throws? form)
(throws? c form)
(throws? c pattern form)
Evals form
and returns true iff it throws an error that matches given
criteria:
c
may be:
:all
=> any platform error (Throwable or js/Error, etc.):common
=> common platform error (Exception or js/Error)pattern
may be:
ex-message
will be matched.ex-data
will be matched.When an error with (nested) causes doesn't match, a match will be attempted against its (nested) causes.
Useful for unit tests, e.g.: (is (throws? {:a :b} (throw (ex-info "Test" {:a :b :c :d}))))
See also throws
.
Evals `form` and returns true iff it throws an error that matches given criteria: - `c` may be: - A predicate function, (fn match? [x]) -> bool - A class (e.g. ArithmeticException, AssertionError, etc.) - `:all` => any platform error (Throwable or js/Error, etc.) - `:common` => common platform error (Exception or js/Error) - `pattern` may be: - A string or Regex against which `ex-message` will be matched. - A map against which `ex-data` will be matched. When an error with (nested) causes doesn't match, a match will be attempted against its (nested) causes. Useful for unit tests, e.g.: (is (throws? {:a :b} (throw (ex-info "Test" {:a :b :c :d})))) See also `throws`.
(time-ms & body)
Returns number of milliseconds it took to execute body.
Returns number of milliseconds it took to execute body.
(time-ns & body)
Returns number of nanoseconds it took to execute body.
Returns number of nanoseconds it took to execute body.
(top n coll)
(top n keyfn coll)
(top n keyfn cmp coll)
Returns a sorted vector of the top n
items from coll
using reduce-top
.
Returns a sorted vector of the top `n` items from `coll` using `reduce-top`.
(top-into to n coll)
(top-into to n keyfn coll)
(top-into to n keyfn cmp coll)
Conjoins the top n
items from coll
into to
using reduce-top
.
Conjoins the top `n` items from `coll` into `to` using `reduce-top`.
(try-eval form)
Evaluates form
. If eval doesn't throw, expands to form
, otherwise to nil.
Evaluates `form`. If eval doesn't throw, expands to `form`, otherwise to nil.
(unexpected-arg! arg & {:keys [msg] :as details})
Throws runtime ExceptionInfo
to indicate an unexpected argument.
Takes optional kvs for merging into exception's data map.
(let [mode :unexpected] (case mode :read (do <...>) :write (do <...>) (unexpected-arg! mode :expected #{:read :write}))) =>
Unexpected argument: :unexpected {:arg {:value :unexpected, :type clojure.lang.Keyword}, :expected #{:read :write}}
Throws runtime `ExceptionInfo` to indicate an unexpected argument. Takes optional kvs for merging into exception's data map. (let [mode :unexpected] (case mode :read (do <...>) :write (do <...>) (unexpected-arg! mode :expected #{:read :write}))) => Unexpected argument: :unexpected {:arg {:value :unexpected, :type clojure.lang.Keyword}, :expected #{:read :write}}
(update-in m ks f)
(update-in m ks not-found f)
Like core/update-in
but:.
not-found
.:swap/dissoc
, :swap/abort
.Like `core/update-in` but:. - Empty ks will return (f m), not act like [nil] ks. - Adds support for `not-found`. - Adds support for special return vals: `:swap/dissoc`, `:swap/abort`.
(url-decode s & [encoding])
Stolen from http://goo.gl/99NSR1
Stolen from http://goo.gl/99NSR1
(url-encode s & [encoding])
(url-encode s)
Based on https://goo.gl/fBqy6e
Based on https://goo.gl/fBqy6e
(uuid-str)
(uuid-str max-length)
Returns a UUIDv4 string of form "xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx". Ref. http://www.ietf.org/rfc/rfc4122.txt, https://gist.github.com/franks42/4159427, https://github.com/clojure/clojurescript/pull/194
Returns a UUIDv4 string of form "xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx". Ref. http://www.ietf.org/rfc/rfc4122.txt, https://gist.github.com/franks42/4159427, https://github.com/clojure/clojurescript/pull/194
(when test-or-bindings & body)
Like core/when
but acts like when-let
when given a binding vector
as test expr.
Like `core/when` but acts like `when-let` when given a binding vector as test expr.
(when-let bindings & body)
Like core/when-let
but can bind multiple values for body
iff all tests
are truthy, supports internal unconditional :let
s.
Like `core/when-let` but can bind multiple values for `body` iff all tests are truthy, supports internal unconditional `:let`s.
(when-not test-or-bindings & body)
Like core/when-not
but acts like when-let
when given a binding vector
as test expr.
Like `core/when-not` but acts like `when-let` when given a binding vector as test expr.
(with-dynamic-assertion-data data & body)
Prefer with-data
Prefer `with-data`
(with-truss-data data & body)
Executes body with dynamic assertion data bound to given value. This data will be included in any violation errors thrown by body.
Executes body with dynamic assertion data bound to given value. This data will be included in any violation errors thrown by body.
cljdoc is a website building & hosting documentation for Clojure/Script libraries
× close