Step 18 of 18 of the SMILES tutorial. One last string, and this time the molecule has a shape. CN1[C@H]2CC[C@@H]1[C@H]([C@H](C2)OC(=O)c3ccccc3)C(=O)OC is cocaine: an N-methyl nitrogen opens ring bond number 1, the atom after it opens ring 2, and the two rings that close on them share the nitrogen and the two carbons beside it — the bridged tropane skeleton — with a benzoate ester hanging in a branch and a methyl ester at the end. Four ring atoms are stereocenters and each carries its own chirality tag; every tag is read on its own, looking from that atom's first neighbour in writing order, so a run of @ and @@ says nothing by itself about which substituents end up on the same face. Delete all four tags: the string still parses, the formula is unchanged, and the depiction loses its wedges — what is left stands for every stereoisomer at once, only one of which is the natural alkaloid. Put them back one at a time and watch the shape come back.
It opens on CN1[C@H]2CC[C@@H]1[C@H]([C@H](C2)OC(=O)c3ccccc3)C(=O)OC, which you can edit in place: the structure follows every keystroke, and so does the notation when you draw instead.
Convert between chemical structures and SMILES in both directions, search a list of structures by substructure or SMARTS, and learn the notation step by step. Everything runs in your browser — no structure you open ever leaves your machine — so the tool needs JavaScript.