diff --git a/README.md b/README.md
index fc2fe65..4dcaf68 100644
--- a/README.md
+++ b/README.md
@@ -1 +1,6 @@
-# strings
\ No newline at end of file
+# A Python FFI provider for purescript-strings
+Providing FFI files for https://github.com/purescript/purescript-strings
+
+Available mirrors:
+
+https://github.com/purescript-python/purescript-python-ffi-index
diff --git a/python-ffi/Data/String/CodePoints.py b/python-ffi/Data/String/CodePoints.py
new file mode 100644
index 0000000..3148580
--- /dev/null
+++ b/python-ffi/Data/String/CodePoints.py
@@ -0,0 +1,73 @@
+def _unsafeCodePointAt0(fallback):
+    def ap(s):
+        return ord(s[0])
+
+    return ap
+
+
+def _codePointAtImpl(fallback, just, nothing, unsafeCodePointAt0, index, s):
+    length = len(s)
+    if index < 0 or index >= length:
+        return nothing
+    else:
+        return just(unsafeCodePointAt0(s[index]))
+
+
+def _codePointAt(fallback):
+    return lambda just: lambda nothing: lambda unsafeCodePointAt0: lambda index: lambda s: _codePointAtImpl(
+        fallback, just, nothing, unsafeCodePointAt0, index, s
+    )
+
+
+def _codePrefixImpl(fallback, unsafeCodePointAt0, pred, s):
+    cpCount = 0
+    for x in s:
+        cp = unsafeCodePointAt0(x)
+        if not pred(cp):
+            return cpCount
+        else:
+            cpCount += 1
+    return cpCount
+
+
+def _fromCodePointArray(singleton):
+    def ap(cps):
+        return "".join(map(_singleton, cps))
+
+    return ap
+
+
+def _singleton(fallback):
+    return chr
+
+
+def _take(fallback):
+    def ap(n):
+        return lambda s: s[:n]
+
+    return ap
+
+
+def _toCodePointArray(fallback):
+    def ap(unsafeCodePointAt0):
+        return lambda s: list(map(unsafeCodePointAt0, s))
+
+    return ap
+
+
+def _countPrefix(fallback):
+    def unsafeCodePointAt0_(unsafeCodePointAt0):
+        def pred_(pred):
+            def str_(s):
+                count = 0
+                for x in s:
+                    if not pred(x):
+                        return count
+                    else:
+                        count += 1
+
+            return str_
+
+        return pred_
+
+    return unsafeCodePointAt0_
diff --git a/python-ffi/Data/String/CodeUnits.py b/python-ffi/Data/String/CodeUnits.py
new file mode 100644
index 0000000..328c749
--- /dev/null
+++ b/python-ffi/Data/String/CodeUnits.py
@@ -0,0 +1,121 @@
+def fromCharArray(a):
+    return "".join(a)
+
+
+def toCharArray(s):
+    return list(s)
+
+
+def singleton(c):
+    return c
+
+
+def _charAt(just):
+    def nothing_(nothing):
+        def i_(i):
+            def s_(s):
+                if i >= 0 and i < len(s):
+                    return just(s[i])
+                else:
+                    return nothing
+
+            return s_
+
+        return i_
+
+    return nothing_
+
+
+def _toChar(just):
+    return lambda nothing: lambda s: just(s) if len(s) == 1 else nothing
+
+
+length = len
+
+
+def countPrefix(p):
+    def ap(s):
+        i = 0
+        while i < len(s) and p(s[i]):
+            i += 1
+        return i
+
+    return ap
+
+
+def _indexOfImpl(just, nothing, x, s):
+    i = s.find(x)
+    return nothing if i == -1 else just(i)
+
+
+def _indexOf(just):
+    return lambda nothing: lambda x: lambda s: _indexOfImpl(just, nothing, x, s)
+
+
+def _indexOfPImpl(just, nothing, x, startAt, s: str):
+    i = s.find(x, startAt)
+    return nothing if i == -1 else just(i)
+
+
+def _indexOfStartingAtImpl(just, nothing, x, startAt, s):
+    if startAt < 0 or startAt > len(s):
+        return nothing
+    i = s.find(x, startAt)
+    return nothing if i == -1 else just(i)
+
+
+globals()[
+    "_indexOf'"
+] = lambda just: lambda nothing: lambda x: lambda startAt: lambda s: _indexOfStartingAtImpl(
+    just, nothing, x, startAt, s
+)
+
+
+def _lastIndexOfImpl(just, nothing, x, s: str):
+    i = s.rfind(x)
+    return nothing if i == -1 else just(i)
+
+
+def _lastIndexOfPImpl(just, nothing, x, startAt, s: str):
+    i = s.rfind(x, startAt)
+    return nothing if i == -1 else just(i)
+
+
+def _lastIndexOf(just):
+    return lambda nothing: lambda x: lambda s: _lastIndexOfImpl(just, nothing, x, s)
+
+
+def _lastIndexOfStartAtImpl(just, nothing, x, startAt, s):
+    if startAt < 0 or startAt > len(s):
+        return nothing
+    i = s.rfind(x, 0, startAt + len(x))
+    return nothing if i == -1 else just(i)
+
+
+globals()[
+    "_lastIndexOf'"
+] = lambda just: lambda nothing: lambda x: lambda startAt: lambda s: _lastIndexOfStartAtImpl(
+    just, nothing, x, startAt, s
+)
+
+
+def _lastIndexOfStartingAt(just):
+    return lambda nothing: lambda x: lambda startAt: lambda s: _lastIndexOfStartAtImpl(
+        just, nothing, x, startAt, s
+    )
+
+
+def take(n: int):
+    return lambda s: s[: max(n, 0)]
+
+
+def drop(n: int):
+    return lambda s: s[max(n, 0) :]
+
+
+def _slice(b):
+    return lambda e: lambda s: s[b:e]
+
+
+def splitAt(i):
+    return lambda s: {"before": s[: max(i, 0)], "after": s[max(i, 0) :]}
diff --git a/python-ffi/Data/String/Common.py b/python-ffi/Data/String/Common.py
new file mode 100644
index 0000000..d760078
--- /dev/null
+++ b/python-ffi/Data/String/Common.py
@@ -0,0 +1,48 @@
+def _localeCompareImpl(lt, eq, gt, s1, s2):
+    # TODO: write correct localeCompare
+
+    # var result = s1.localeCompare(s2);
+    # return result < 0 ? lt : result > 0 ? gt : eq;
+    if s1 == s2:
+        return eq
+    if s1 < s2:
+        return lt
+    if s1 > s2:
+        return gt
+
+
+def _localeCompare(lt):
+    return lambda eq: lambda gt: lambda s1: lambda s2: _localeCompareImpl(
+        lt, eq, gt, s1, s2
+    )
+
+
+def replace(s1):
+    return lambda s2: lambda s3: s3.replace(s1, s2, 1)
+
+
+def replaceAll(s1):
+    return lambda s2: lambda s3: s3.replace(s1, s2)
+
+
+def split(sep):
+    if len(sep) == 0:
+        return lambda s: list(s)
+    else:
+        return lambda s: s.split(sep)
+
+
+def toLower(s):
+    return s.lower()
+
+
+def toUpper(s):
+    return s.upper()
+
+
+def trim(s):
+    return s.strip()
+
+
+def joinWith(s):
+    return lambda xs: s.join(xs)
diff --git a/python-ffi/Data/String/Regex.py b/python-ffi/Data/String/Regex.py
new file mode 100644
index 0000000..4646046
--- /dev/null
+++ b/python-ffi/Data/String/Regex.py
@@ -0,0 +1,110 @@
+import re
+
+
+def showRegexImpl(r):
+    return "" + r
+
+
+globals()["showRegex'"] = showRegexImpl
+
+
+def regexImpl(left):
+    raise NotImplementedError("FFI Not implemented. Data.String.Regex.regexImpl")
+
+
+#   return function (right) {
+#     return function (s1) {
+#       return function (s2) {
+#         try {
+#           return right(new RegExp(s1, s2));
+#         } catch (e) {
+#           return left(e.message);
+#         }
+
+globals()["regex'"] = regexImpl
+
+
+def source(r):
+    return r["source"]
+
+
+def flagsImpl(r):
+    return {
+        "multiline": r["multiline"],
+        "ignoreCase": r["ignoreCase"],
+        "global": r["global"],
+        "sticky": bool(r["sticky"]),
+        "unicode": bool(r["unicode"]),
+    }
+
+
+globals()["flags'"] = regexImpl
+
+
+def test(r):
+    raise NotImplementedError("FFI Not implemented. Data.String.Regex.test")
+
+
+#   return function (s) {
+#     var lastIndex = r.lastIndex;
+#     var result = r.test(s);
+#     r.lastIndex = lastIndex;
+#     return result;
+
+
+def _matchImpl(just, nothing, r, s):
+    m = re.match(r, s)
+    if m is None:
+        return nothing
+    else:
+        return just(m)
+
+
+_match = lambda j: lambda n: lambda r: lambda s: _matchImpl(j, n, r, s)
+
+
+#   return function (nothing) {
+#     return function (r) {
+#       return function (s) {
+#         var m = s.match(r);
+#         if (m == null || m.length === 0) {
+#           return nothing;
+#         } else {
+#           for (var i = 0; i < m.length; i++) {
+#             m[i] = m[i] == null ? nothing : just(m[i]);
+#           }
+#           return just(m);
+#         }
+
+
+def replace(r):
+    return lambda s1: lambda s2: s2.replace(s1, 1)
+
+
+def replaceBy(r):
+    raise NotImplementedError("FFI Not implemented. Data.String.Regex.replaceBy")
+
+
+globals()["replace'"] = replaceBy
+
+#   return function (f) {
+#     return function (s2) {
+#       return s2.replace(r, function (match) {
+#         return f(match)(Array.prototype.splice.call(arguments, 1, arguments.length - 3));
+#       });
+
+
+def _searchImpl(just, nothing, r, s):
+    m = re.search(r, s)
+    if m:
+        return just(m.start())
+    else:
+        return nothing
+
+
+def _search(just):
+    return lambda nothing: lambda r: lambda s: _searchImpl(just, nothing, r, s)
+
+
+def split(r):
+    return lambda s: s.split(r)
diff --git a/python-ffi/Data/String/Unsafe.py b/python-ffi/Data/String/Unsafe.py
new file mode 100644
index 0000000..455d7e8
--- /dev/null
+++ b/python-ffi/Data/String/Unsafe.py
@@ -0,0 +1,14 @@
+def charAt(i):
+    def ap(s):
+        if i >= 0 and i < len(s):
+            return s[i]
+
+        raise IndexError("Data.String.Unsafe.charAt: Invalid index.")
+
+    return ap
+
+
+def char(s):
+    if len(s) == 1:
+        return s[0]
+    raise ValueError("Data.String.Unsafe.char: Expected string of length 1.")